CV-20-00643629-00CL
ONTARIO
SUPERIOR COURT OF JUSTICE
(Commercial List)
BETWEEN:
FAIRSTONE FINANCIAL HOLDINGS INC., J.C. FLOWERS IV L.P. and VP CANADA ACQUISITION, LP
Applicants
- and –
DUO BANK OF CANADA
Respondent
AND BETWEEN:
DUO BANK OF CANADA
Plaintiff
- and –
FAIRSTONE FINANCIAL HOLDINGS INC., J.C. FLOWERS IV COINVEST CANADA L.P., J.C. FLOWERS IV L.P., and VP CANADA ACQUISITION, LP
Defendants
Eliot N. Kolers, Danielle K. Royal, Sinziana R. Hennig, Zev Smith and Alexander DeParde for the Applicants
Linda Plumpton, Andrew Gray, Gillian Dingle, Leora Jackson, Adrienne Oake and Alicja Puchta for the Respondent
HEARD: September 8-11, 14 and 15, 2020
Koehnen, J.
Overview
1On February 18, 2020, Fairstone Financial Holdings Inc. and its subsidiaries signed a share purchase agreement with the respondent Duo Bank of Canada pursuant to which Duo agreed to purchase Fairstone’s business. Closing was scheduled for June 1, 2020. On May 27, 2020 Duo advised Fairstone that it would not be closing on June 1 as planned. Duo took the position that four covenants in the share purchase agreement allowed it to avoid closing: the material adverse event covenant, the ordinary course covenant, the amortization event covenant and the access to information covenant.
2The first three covenants are allegedly triggered by events that occurred or were expected to occur as a result of the Covid -19 pandemic which the World Health Organization declared on March 11, 2020.
3In response, Fairstone brought this application for specific performance. If I find against it, Duo prefers specific performance over damages as the appropriate remedy.
4For the reasons set out below, I find in favour of Fairstone and order Duo to specifically perform the share purchase agreement.
5Fairstone covenanted in the share purchase agreement, that no material adverse effect shall have occurred between the date of signing and the date of closing. At first blush, I am prepared to agree that a material adverse effect did occur as a result of the pandemic. There are, however, three carveouts from the material adverse effect clause that apply here to take Duo’s complaints outside of the scope of the clause.
6Material adverse effect is defined so as to exclude effects that are caused by (i) worldwide, national, provincial or local emergencies; (ii) changes in the markets or industry in which Fairstone operates; or (iii) the failure of Fairstone to meet any financial projections. Here the adverse effect was clearly caused by the pandemic which falls into the definition of the first carveout. In addition, the changes of which Duo complains are changes to the entire market and industry in which Fairstone operates. They are not changes unique to Fairstone. Third, Duo’s fundamental complaint is that Fairstone failed to meet the projections contained in its financial plan. The first two carveouts have the added nuance that they apply only to the extent that Fairstone has not been disproportionately adversely affected relative to other persons in the industries or markets in which Fairstone operates. I have concluded that Fairstone has not been disproportionately affected.
7The ordinary course covenant requires Fairstone to act between signing and closing in a manner that is consistent with its past practices and that is in the ordinary course of normal day-to-day operations, to the extent it is lawfully able to do so. Fairstone acted in a manner that was consistent with its past practice. The steps Fairstone put into place after the pandemic was declared were consistent with steps Fairstone had put into place during past recessions and/or were steps that one would expect a business to put into place as part of its ordinary course operations during a recession. To the extent that Fairstone took steps that it had not taken in the past, it was required to do so by law. Courts have developed a number of principles to help determine whether conduct is within the ordinary course. Fairstone’s behaviour falls well within those principles.
8The amortization event covenant is one pursuant to which Fairstone agreed, among other things, that its losses within its loan portfolios would not exceed a certain trigger point. All agree that the losses did not exceed that trigger point before the end of trial. The issue turns on whether, in the language of the agreement, a circumstance exists that “would reasonably be expected to result in an amortization event .” The short answer to Duo’s complaint is that simply because a financial model might project an amortization event does not mean that it will occur. The whole point of obtaining those projections is to enable management to recalibrate the business to avoid the projected amortization event from ever occurring. Duo had a high level of confidence in Fairstone’s management. Indeed, one of the attractions of Fairstone for Duo was that Fairstone had proprietary analytics and a series of management teams, including current management, that had successfully steered Fairstone through recessions for approximately 100 years.
9The access to information covenant is one that required Fairstone to furnish Duo with such financial and operating data as was reasonably necessary for Duo to consummate the transaction. Duo did not rely on any alleged breach of this covenant when it announced that it would not close the transaction on June 1. Fairstone responded to thousands of information requests. By the time of trial, Duo alleges that 9 were unanswered and 7 were insufficiently answered. In my view, the requests amounted to more of a fishing expedition in which Duo was looking for reasons not to close as opposed to information requests that were necessary to consummate the transaction.
I. The Parties and the Transaction
10Fairstone Financial Holdings Inc. and its subsidiaries constitute Canada’s largest consumer finance company that targets near prime borrowers. It has operated since 1923, and has over $3 billion in assets.
11It is currently owned by J.C. Flowers and Värde Partners. Flowers is a U.S.-based private investment firm with approximately $6 billion of assets under management. Värde is a U.S.-based global alternative investment firm with approximately $14 billion of assets under management. In late 2019, Flowers and Värde put Fairstone up for sale through a competitive bidding process.
12Duo Bank of Canada participated in the bidding. Duo Bank is a privately-owned Schedule I Canadian Bank. Its products and services include credit cards and related protection services deposits and reward programs. Duo Bank had tried to purchase Fairstone when it was last up for sale but lost out to Flowers and Värde.
13The bidding process was highly competitive. Duo increased its initial offer for the Company significantly over a short interval.1 On February 18, 2020, Fairstone and Duo signed a share purchase agreement(the “SPA”) pursuant to which Duo agreed to purchase the shares of Fairstone for its tangible shareholders’ equity plus very substantial premium. While the precise sales price would depend on the tangible shareholders’ equity on closing, it was estimated to be over $1 billion.
14The purchase price was funded by Centerbridge Partners L.P., the Ontario Teachers’ Pension Plan Board and Stephen Smith. Centerbridge is a private investment firm that manages approximately $27 billion in capital in New York and London. It has approximately $20 billion invested in financial services, consumer credit and related assets including in Duo Bank. The Ontario Teachers Pension Plan is one of Canada’s largest and highest profile pension funds with over $200 billion in assets under management. Stephen Smith is a highly experienced entrepreneur in the financial services industry. He is Chair, CEO and that of First National Financial Corporation, a Canadian-based originator and underwriter of mortgages with over $110 billion under administration. He is also a director of the Canada Infrastructure Bank and of E - L Financial Corporation.
15There is no doubt that all parties to the transaction at issue were on the high end of the sophistication spectrum. All were experienced deal makers who were assisted by sophisticated legal and financial advisors.
16The parties anticipated a closing on June 1, 2020, but provided for an extended closing up to August 14, 2020.
17The SPA had relatively minimal conditions. Fairstone characterizes this as the result of Duo having failed during the last bidding process and wanting to ensure it succeeded this time. Duo’s bid promised the “highest level of certainty in terms of both execution and timing” as a key feature of its bid. In addition, Fairstone says Centerbridge had a reputation for retrading issues that had already been settled. As a result, says Fairstone, it wanted to keep closing conditions to a minimum.
18Duo agrees that the number of closing conditions was small but characterizes them as being of pivotal importance because they were so few in number. Duo characterizes the closing conditions as protecting it
…against adverse changes to the business it had agreed to buy. These protections obligated Fairstone to maintain its business in the short period between signing and closing, guarded against adverse outcomes resulting from the looming COVID – 19 pandemic, and ensured that by the time the transaction closed, no major deterioration in Fairstone’s financing facilities or the underlying collateral would have occurred.
19As the Covid pandemic evolved, so did tensions between the parties. On April 1, 2020, Duo took the position that Fairstone may already be in breach of the Material Adverse Effect, Ordinary Course and Amortization Event covenants in the SPA.
20Tensions continued to mount and culminated in a letter from Duo to Fairstone dated May 27, 2020, in which Duo stated that it did not intend to close the transaction on June 1 as planned. In response, Fairstone initiated this litigation. Duo commenced a parallel proceeding.
21Each side accuses the other of buyer’s or seller’s remorse, wishing that it could have negotiated an agreement with better protections for its position.
22Fairstone takes the position that it is entitled to specific performance of the SPA. Duo takes the position that is entitled to terminate the transaction. Which side prevails turns on the determination of the following issues:
(i) Has there been a Material Adverse Effect?
(ii) Has Fairstone breached the Ordinary Course Covenant?
(iii) Has an Amortization Event occurred or is one reasonably likely to occur?
(iv) Has a material breach of the Information Covenant occurred?
II. Material Adverse Effect
23Section 6.2 of the SPA contains closing conditions for the benefit of the purchasers. Section 6.2 (2) provides:
No Material Adverse Effect. Between the date of this Agreement and the Effective Time2, there shall not have occurred a Material Adverse Effect.
24Material adverse effect is defined in section 1.1 (ccc) as follows:
“Material Adverse Effect” means a fact, circumstance, condition, change, event or occurrence that has (or would reasonably be expected to have), individually or in the aggregate, a material adverse effect on the Business, operations, assets, liabilities or condition (financial or otherwise) of the Acquired Companies, taken as a whole; except to the extent that the material adverse effect results from or is caused by:
(i) worldwide, national, provincial or local conditions or circumstances, whether they are economic, political, regulatory (including any change in Law or IFRS) or otherwise, including war, armed hostilities, acts of terrorism, emergencies, crises and natural disasters;
(ii) changes in the markets or industry in which the Acquired Companies operate;
(vi) the failure of any of the Acquired Companies to meet any internal, published or public projections, forecasts, guidance or estimates, including without limitation of production, revenues, earnings or cash flows (it being understood that the causes underlying such failure may be taken into account in determining whether a Material Adverse Effect has occurred);
provided, however, that in the case of the events described in each of the clauses (i) and (ii), that such events do not have a materially disproportionate adverse impact on the Acquired Companies relative to other Persons in the industries or markets in which the Acquired Companies operate.
25Like other contractual clauses, material adverse effect (“MAE”)3 clauses are to be interpreted in light of their purpose and object.
26The overall purpose of a MAE clause is to allocate risks between the seller and the purchaser. Generally speaking, MAE clauses protect purchasers from developments that will cause a target company to be materially different at closing from what it was when the agreement was entered into.4
27The MAE at issue here begins by having the seller covenant that no Material Adverse Effects will occur between the date of the agreement and closing.5 It then carves out seven events which do not constitute a MAE even though they otherwise might. Three carveouts are relevant here: the carveout for MAEs caused by emergencies, those caused by changes to the markets or industry in which Fairstone operates and those caused by Fairstone’s failure to meet financial projections.
28Thus, if there is a MAE that is caused by an emergency, a change in Fairstone’s industry/market or Fairstone’s failure to meet financial projections, Duo must, prima facie, complete the purchase.
29There is, however, an exception to the emergency and market change carveouts. If the MAE is caused by one of those two conditions, Duo must complete the purchase unless the emergency or market change has had a materially disproportionate adverse impact on Fairstone compared to other persons in the industries or markets in which Fairstone operates. If the emergency or market change has had such an effect, then Duo need not complete the purchase after all.
30The MAE clause in the SPA gives rise to the following issues between the parties:
A. The burden of proof.
B. Timing issues.
C. The definition of material adverse effect.
D. The proper interpretation of the carveouts and the exceptions to the carveouts in subparagraphs (i), (ii) and (vi) of s. 1.1 (ccc).
A. Burden of Proof
31It appears that Ontario courts have not yet expressed a view about which party bears the burden of proof to establish that a MAEhas occurred. American, English and Quebec cases have all held that the burden of proof falls on the party asserting that a MAE has occurred.6 This makes good sense, particularly given that the party asserting the MAE is trying to use that allegation to avoid a contract they would otherwise be obliged to perform.
32Duo agrees that it bears the burden of establishing a MAE but submits that the burden shifts back to Fairstone to establish that one of the carveouts to the definition of MAE is present. I accept that submission. It is consistent with the general proposition that the burden of proof with respect to any particular proposition lies on the party asserting it.7
33I would, however, go one step further and shift the burden of proof back to Duo if it wants to argue that the carveout does not apply because Fairstone has, to use the language of the SPA, suffered a “materially disproportionate adverse impact… relative to other Persons in the industries or markets in which” Fairstone operates. I note that this is also the approach that the Delaware Court of Chancery took in Akorn, Inc. v. Fresenius Kabi AG.8
34While a shifting burden of proof may appear complex at first, it is really nothing more than a reflection of the principle that a party who makes an assertion, bears the onus of proving it.
35A further layer of complication might arise because section 1.1 (ccc) of the SPA defines MAE as a condition:
…that has (or would reasonably be expected to have)… a material adverse effect on the Business…
36What then is the appropriate burden of proof for the assertion that something “would reasonably be expected to have” a material adverse effect?
37Duo submits that this requires proof of the expected event at a standard lower than the civil burden of proof. Fairstone, on the other hand, submits that the burden of proof required is the ordinary balance of probabilities test. In support of this proposition, Fairstone relies on authorities to the effect that there is only one civil burden of proof.9
38The Supreme Court of Canada has come to potentially conflicting views on the question in the limited jurisprudence in which it has considered the issue when interpreting similar language in statutes.
39In Merck Frosst Canada Ltd. v. Canada (Health),10 the Supreme Court of Canada dealt with a freedom of information request. The statute allowed documents in response to such requests to be redacted if there was a “reasonable expectation of probable harm” to a third party. In considering that test, the Supreme Court of Canada held that:
A “reasonable expectation” is something that is at least foreseen and perhaps likely to occur, but not necessarily probable. When the two expressions are used in combination — “a reasonable expectation of probable harm” — the resulting standard is perhaps not immediately apparent. However, I conclude that this long-accepted formulation is intended to capture an important point: while the third party need not show on a balance of probabilities that the harm will in fact come to pass if the records are disclosed, the third party must nonetheless do more than show that such harm is simply possible. Understood in that way, I see no reason to reformulate the way the test has been expressed. (emphasis added)11
This articulation sounds like something less than the ordinary civil burden of proof.
40In Hilewitz v. Canada (Minister of Citizenship and Immigration),12 the Supreme Court of Canada interpreted the statutory phrase “would cause or might reasonably be expected to cause excessive demands on health or social services.” In describing the burden of proof associated with that phrase, the court stated:
The threshold is reasonable probability, not remote possibility. It should be more likely than not… that the contingencies will materialize.”13 (emphasis added)
This articulation sounds more like the ordinary civil burden of proof. The Court also emphasized that anticipated burdens must be “tethered to the realities, not the possibilities.”14
41American cases have noted that future occurrences can qualify as MAEs and that a MAE can have occurred without the effect having been felt yet15 although a “mere risk of a MAE cannot be enough”.16
42American jurisprudence also notes that the word “would” in the expression “would reasonably be expected to have a MAE” suggests a higher degree of likelihood than words such as could or might17 and conclude that the purchaser must show that the MAE is “more likely than not” to occur. 18 This too sounds more like the ordinary civil burden of proof.
43When speaking about contractual language similar to that in the SPA, the court in Akorn noted:
Even under this standard, a mere risk of an MAE cannot be enough. "There must be some showing that there is a basis in law and in fact for the serious adverse consequences prophesied by the party claiming the MAE." When evaluating whether a particular issue would reasonably be expected to result in an MAE, the court must consider "quantitative and qualitative aspects." "It is possible, in the right case, for a party . . . to come forward with factual and opinion testimony that would provide a court with the basis to make a reasonable and an informed judgment of the probability of an outcome on the merits."19 (Citations omitted)
44From the foregoing discussion I glean the following principles: Duo must show on a balance of probabilities that the condition etc. at issue would reasonably be expected to have a MAE on Fairstone. In doing so, Duo need not demonstrate on a balance of probabilities that the condition will have a MAE; rather, the balance of probabilities relates to the condition being reasonably expected to have such an effect. In determining whether a condition is reasonably expected to have a MAE, Duo must show:
more than a possibility or risk;
evidence that allows the court to reach an informed judgment;
that the evidence the court relies on and the judgment it forms must be tethered to realities and not to mere possibilities;
the evidence relied on can be quantitative and/or qualitative.
45I will return to these principles when assessing whether a MAE has occurred, whether a carveout applies and whether Fairstone has been disproportionately affected by any carveout.
B. Timing Issues
46Two timing issues arise with respect to the interpretation of the MAE clause:
i. As of what date does one assess whether a MAE has arisen?
ii. How far into the future can one look to determine if a condition is reasonably expected to have a MAE on the business?
i. Date of Assessment
47Duo submits that the date on which to assess whether a MAE has arisen is May 27, 2020, the date on which Duo advised that it would not close the transaction on June 1 as planned. Fairstone, on the other hand, submits that August 14, 2020, is the date on which to ask the question because that was the outside date by which the transaction had to close under the terms of the SPA.
48Each side perceives its chosen date as being more advantageous to itself. Duo believes that Fairstone’s reported financial condition was poorer on May 27, 2020, than it was on August 14. This is so because in April 2020 Fairstone issued a revised forecast that projected it would be taking a loan loss reserve of $35 million in the second quarter. Had this been done, Fairstone would have incurred a not insignificant loss for that quarter. As a result, in Duo’s view, Fairstone had suffered a MAE. By the end of June, however, things had improved materially so that Fairstone did not take the projected loan loss reserve and did not suffer the projected loss of the second quarter.
49Duo contests the validity of the adjustments Fairstone made to its projections. That issue will be addressed later in these reasons.
50In support of its submission, Duo cites a number cases that refer to the date of termination as the appropriate time to assess whether a MAE has occurred because that is the date on which the purchaser took action in response to the alleged MAE.20
51In my view, August 14, 2020, is the preferable date on which to determine whether a MAE has occurred or is reasonably expected to occur. Although I accept the authorities that Duo relies on, they are distinguishable. They involved situations in which the purchaser actually terminated the transaction because of a MAE.
52Duo did not, however, terminate the SPA on May 27 but merely stated:
Purchaser is not required, and does not intend, to close on June 1, 2020.
53The omission of language terminating the SPA was not accidental. Throughout the trial, Fairstone highlighted the fact that Duo never terminated the SPA. Duo never took issue with that assertion.
54It made good sense for Duo not to terminate. Duo was the successful and aggressive bidder in an auction. Duo knew that if it terminated, Fairstone would not be able to sell the business for the same price that Duo had offered. If Duo turned out to be wrong about its right to terminate, it would be responsible for damages potentially in the hundreds of millions of dollars. Duo’s position at trial supports this analysis. While Duo maintained that it was entitled not to close, it also made clear that it preferred specific performance over damages as a remedy.
55Using August 14 is additionally appropriate because, by failing to terminate the SPA, Duo reserved to itself the ability to close as late as August 14.21 Having reserved that advantage to itself, it must also accept any perceived attendant disadvantages. Moreover, given that the absence of a MAE is a closing condition, it is appropriate to use the closing date, actual or conceptual, as the date on which to ask the question.
ii. How Far into the Future Should the Court Look?
56As noted earlier, the MAE clause speaks of a condition “that has (or would reasonably be expected to have)” a material adverse effect on the business. The parenthetical language is prospective which suggests that the adverse effect need not yet have occurred.
57How far into the future should courts look to determine whether something would reasonably be expected to have a MAE?
58On the one hand, the forward-looking period cannot be indefinite. At some point most businesses are bound to encounter a MAE whether caused by external economic factors like a recession or by internal factors like a management error. As a general rule, a purchaser takes on forward-looking risks when it acquires a business. A seller, even with a MAE drafted like the one in the SPA, should not be expected to protect the purchaser against all future economic dislocations.
59Duo, citing language from Akorn, submits that the correct analysis of this language in relation to the MAE is to look out “over a commercially reasonable period, which one would expect to be measured in years rather than months.”22 Akorn, however, uses that language not to determine how far into the future one should look to determine whether a MAE would reasonably be expected to arise but in reference to how long a particular condition must last for it to be considered a MAE.
60How far into the future courts should look to see if a MAE is expected to arise will inevitably depend on the circumstances of the case. If a purchaser is acquiring a cyclical business and by the time of closing can foresee a recession within two years of closing, it would not be appropriate to allow the purchaser to avoid the transaction. That is the sort of longer-term cyclical business risk a purchaser assumes when it acquires a business. Allowing a purchaser to avoid a transaction for something of that sort would encourage potential purchasers to make aggressive bids for businesses and then take the position at closing that a recession was in the offing within the next two years thereby letting them renegotiate the purchase or refuse to close.
61On the other hand, one can envisage circumstances that arise between signing and closing that would not have a MAE on the acquired business until months after closing yet would warrant allowing the purchaser to avoid the transaction. This could include things such as the emergence of a disruptive technology, notice that a key customer will not renew its contract, notice that a license will not be renewed at the end of its term and threats to a business’ ability to use intellectual property.
62I will return to these principles when assessing whether a MAE has occurred, whether a carveout applies and whether Fairstone has been disproportionately affected by any carveout.
C. Definition of MAE
63The definition of a MAE in the SPA is not particularly helpful because it refers back to itself. The critical wording of section 1.1 (ccc) defines MAE to mean a
… condition that has (or would reasonably be expected to have) … a material adverse effect on the Business…”
64More helpful is a widely used definition in American jurisprudence that defines MAE as:
…the occurrence of unknown events that substantially threaten the overall earnings potential of the target in a durationally-significant manner.23
65That definition contains three elements: an unknown event, a threat to overall earnings potential and durational significance. For the reasons set out below, I conclude that Duo has met its burden of proof to establish that a MAE arose between signing and both May 27 and August 14, 2020.
i. Unknown Event
66Although I have cited the unknown element of a MAE to American authority, the requirement for the unknown nature of the event has also been adopted in Canadian case law.24 As the British Columbia Supreme Court noted in Inmet Mining Corp. v. Homestake Canada Inc.:
In my view, the key to determining materiality and adversity in the circumstances of the case before me is the knowledge the defendant had as purchaser. If a fact or information were already known to the defendant, or if the defendant did not rely on it, the failure of the plaintiff to disclose it or information related to it would be of no consequence to the defendant's decision to buy and therefore would not be material or adverse to the defendant.25
67Duo admits in its written argument that:
- By the time the parties entered into the SPA, the novel coronavirus was already daily news in North America, and the likelihood of a pandemic was apparent.
68In light of that admission, one could argue that Covid–19 was a known event and can therefore not constitute a MAE.
69There are, however, several countervailing considerations. The contract here does not speak of an unknown condition but simply refers to a condition.
70The language of the MAE clause at issue says that a MAE shall not have occurred between signing and closing. It then defines MAE as a “condition etc. that has (or would reasonably be expected to have) a material adverse effect on the business…” The effect is therefore integral to the definition.
71Although the virus may have been known to both parties when the SPA was signed, there is no evidence before me that either party appreciated the effect that the virus posed for Fairstone.
72Given that MAE clauses are interpreted from the perspective of the party for whose benefit the MAE was granted26 I would be inclined to grant Duo the benefit of the doubt on this issue and find that Covid – 19 satisfies this requirement because its effect on the busines was unknown.
ii. Threat to Overall Earnings Potential
73I am prepared to accept that the Covid–19 pandemic poses a threat to Fairstone’s earnings potential.
74Year-over-year new loan origination had decreased by 56% in May 2020, 36.6% in June 2020 and 21% in July 2020. Fairstone and its experts admitted throughout the trial that one of the tools they use to manage risk in times of economic contraction is to reduce lending and tighten lending requirements. By both May 27 and August 14, 2020, Covid had a material adverse effect on Fairstone’s earnings potential. As of August 14, 2020, the threat to earnings potential would reasonably be expected to continue at least until a vaccine or easily accessible and effective therapeutic treatment was made widely available.
75In addition, Duo points to one of Fairstone’s securitized loan facilities known as the 2019 - 2 facility which consisted of approximately $1.377 billion in outstanding loans. Duo points to a sudden spike in credit losses in that facility during April, May and June 2020 as evidence of a MAE.
76I do not accept that the increase in credit losses in the 2019 – 2 facility constitutes a MAE or any other sort of financial weakening in that facility. Rather, the increase is due to what is known as seasoning of the portfolio and was entirely expected, even by Duo. The spike arises because the facility came into being in October 2019. As a matter of course, Fairstone does not record credit losses in a loan portfolio during its first six months. It then records those losses as a three-month rolling average. The effect of this is that during the first six months of the portfolio’s life, credit losses are either zero or close to zero.27 In the sixth month (in this case, April) there will be a sudden spike because Fairstone has begun recording credit losses. However, the loss recorded in month six will be relatively low because it will be based on a rolling three-month average in which the first two of the three months have losses at zero or close to zero. Month seven will see a further increase in losses because the three-month rolling average will be comprised of two months with recorded losses and one month without recorded losses. Month eight will see a further spike because it now records three full months of losses. In month nine and following, however, the loss curve will flatten because Fairstone will now be in a pattern of recording losses based on a three-month rolling average in which each month contains recorded losses. Duo’s CEO acknowledged on cross-examination that she was aware of this phenomenon and expected the spike and subsequent leveling to occur. In those circumstances the increase in credit losses in the 2019 - 2 facility cannot amount to a MAE.
iii. Durational Significance
77Although Canadian cases have not expressly articulated a durational requirement, it is implicit in their decisions. By way of example, in Mull v. Dynacare Inc.,28 the court found that a drop in income constituted a MAE in part because the decline arose from a policy change by a provincial ministry of health, the only party to whom the business could legally sell services. The loss of income was therefore implicitly permanent.
78American cases have defined durational significance as changes that are “expected to persist significantly into the future.” 29 A short-term drop in earnings does not suffice. Instead, the change must be material when viewed from the longer-term perspective of a reasonable acquirer.30 The length of the durational requirement depends on the context. For a short-term speculator, the durational requirement may be relatively short to constitute a MAE.31
79In other cases, courts have demanded adverse changes in the target's business that are “consequential to the company's long-term earnings power over a commercially reasonable period, which one would expect to be measured in years rather than months.”32
80In IBP Shareholders Litigation v. Tyson Foods, for instance, the target was a beef producer. The Delaware court found that a 64% drop in quarterly earnings did not rise to the level of a MAE because the decline was caused by well-known cyclical shifts in the meat industry, made worse by a harsh winter. Once those two temporary events passed, financial results returned to their historic norms.33
81Duo points to projections that show the pandemic having potential adverse effects on Fairstone into 2022. Is that durationally significant? On the one hand Duo was purchasing for the long term. Duo saw the Fairstone acquisition as an opportunity to gain synergies that would decrease Fairstone’s cost of funds and increase profits of the combined businesses. Duo noted that the acquisition would add scale and diversification to the bank. The synergies were of a nature that they made sense only in the context of a long term investment.34
82Duo’s due diligence materials specifically highlighted Fairstone’s 100-year history of navigating through a variety of economic cycles and that Fairstone’s current, strong, experienced management team had successfully steered Fairstone through the financial crisis of 2008 – 2010.
83There was no evidence led to suggest that Covid-19 will affect the long-term need for consumer finance. The only evidence led in this regard is that, if anything, Covid-19 might help Fairstone in the long run. This can arise because banks tend to tighten their credit underwriting during economic downturns. As a result, some consumers who once qualified for bank financing may no longer do so. Those consumers then turn to lenders one step down from banks; lenders like Fairstone.
84The foregoing factors might lead one to find that the adverse effect that is not durationally significant. On the other hand, projections still show an effect of approximately two years duration. As of August 14, 2020, no vaccine had been approved for release and no therapeutic treatment existed. Even when a vaccine is approved, current estimates project that it will take until the end of 2021 to have widespread inoculation within the population. These eventualities take me back to the parenthetical language that there be no condition, etc., that “would reasonably be expected to have a material adverse effect on the Business”.
85There is certainly more than a possibility or a risk that Covid-19 will have a durationally significant impact on Fairstone. Projections exist that show adverse impacts into 2022. Those projections are more than a possibility or a risk. They amount to evidence that allows for an informed judgment that there is a condition that would reasonably be expected to have a material adverse effect on Fairstone for approximately two years. Projections are unpredictable. The adverse impact could last for a shorter period, it could last longer.
86I return to the principle that MAE clauses are to be interpreted from the perspective of the party for whose benefit they were granted. 35 Here the MAE was clearly granted for Duo’s benefit. This gave rise to considerable debate at trial about the degree to which approaching the question from the perspective of the purchaser required a subjective or an objective test. Needless confusion can arise by using the labels objective or subjective. The test, however, remains clear: one must determine whether the event would affect the decision of a reasonable investor in the circumstances of the purchaser with the information available to it. The subjective views of the purchaser about whether something was a MAE are irrelevant.36
87What then would a reasonable purchaser do in circumstances where, between the signing of the contract and closing, a condition etc. had arisen that had resulted in a decrease of between 21% and 56% in new loan origination which would reasonably be expected to continue into 2022, at levels that are currently unknown. In my view, it would not be unreasonable for a purchaser in those circumstances to try to avoid the transaction, perhaps in the hope of renegotiating at a lower price. That is not to say that all purchasers would do so. A reasonable purchaser could equally decide to close the transaction. I cannot say, however, that it would be objectively unreasonable for purchaser to do anything but seek an exit. I therefore conclude that Duo has met the burden of proof to demonstrate that a MAE arose between signing and both May 27 and August 14, 2020.
D. Carveouts to MAEs
88As set out earlier, after defining material adverse effect, section 1.1 (ccc) of the SPA sets out seven carveouts from the definition. Three are relevant here: the projections miss carveout, the emergency carveout and the market change carveout. Fairstone bears the evidentiary burden of establishing that any of the carveouts apply. I am satisfied that Fairstone has done so in respect of all three carveouts. All three carveouts apply to remove the complaints Duo raises from the ambit of the MAE provisions as a result of which Duo cannot avoid the SPA by asserting a MAE.
i. The Projections Miss Carveout
89I will address the carveouts in an order slightly different from the order in which they appear in s. 1(ccc). I address the projections miss carveout first because it disposes of the MAE issue on its own and is not subject to the disproportionate effect clause.
90I reproduce s. 1 (ccc) (vi) here for convenience. Under that provision, a condition. etc., that otherwise causes a MAE does not constitute a MAE:
…to the extent that the material adverse effect results from or is caused by:
(vi) the failure of any of the Acquired Companies to meet any internal, published or public projections, forecasts, guidance or estimates, including without limitation of production, revenues, earnings or cash flows (it being understood that the causes
underlying such failure may be taken into account in determining whether a Material Adverse Effect has occurred).
91The gist of the MAE that Duo complains of is what it perceives as a downturn in Fairstone’s financial performance due to the Covid-19 pandemic. Its experts, witnesses and written argument go to great lengths to compare performance benchmarks in Fairstone’s Monthly Business Reviews to Fairstone’s Annual Operating Plan and conclude that a shortfall from the plan amounts to a MAE. Even Duo’s longer-term projections of Fairstone’s performance into 2022 are based on comparisons to projections Fairstone published in 2019.
92Carveout (vi) is a seller-friendly provision. One development that purchasers commonly worry about is the failure of the target to maintain budgeted earnings. Section 1(ccc)(vi) specifically excludes the failure to meet financial targets from the ambit of a MAE. Although it contains an exception in that the cause underlying the failure to achieve financial targets can amount to a MAE, the cause of any shortfall from Fairstone’s financial targets is the Covid-19 pandemic. That, as we will see below is barred by the emergency carveout in s. 1(ccc) (i).
93A purchaser who was concerned about shortfalls from targets has a number of mechanisms to protect itself. It could, for example, leave out the projections miss carveout entirely. It could define a MAE as a deterioration in specified financial measures by a specified amount. It could structure part of the purchase price as an earn out contingent on maintaining specific financial benchmarks. One of the factors that Canadian and American courts have identified as relevant to interpreting MAE clauses is the identity of the parties.37 As noted earlier, the parties here were on the highly sophisticated end of the spectrum. All had highly sophisticated advisors. The evidence before me indicated that the MAE provisions were negotiated back and forth between the parties. Given the experience of Duo and its investors, they would be well aware of those and other measures. They employed none of them.
94Vice Chancellor Lamb addressed a similar exclusion in Hexion,38 where the purchaser faced a target that had suffered a loss of 32% of its EBITDA. He found that the projections miss exclusion in that agreement allocated to the purchaser the risk that the target would not meet expectations. As he noted:
Creative investment bankers and deal lawyers could have structured, at the agreement of the parties, any number of potential terms to shift to Huntsman [the target] some or all of the risk that Huntsman would fail to hit its forecast targets. But none of these things happened. Instead, Hexion agreed that the contract contained no representation or warranty with respect to Huntsman’s forecasts. To allow the MAE analysis to hinge on Huntsman’s failure to hit its forecast targets during the period leading up to closing would eviscerate, if not render altogether void, the meaning of section 5.11(b) [the projections miss exclusion].39
95The identical analysis applies here.
96As many courts have noted, MAE clauses must be construed so as to give purchasers the protections they bargained for.40 They should not be stretched to provide either party protections they could have had but did not bargain for.
97Duo appears to recognize the challenges that the projections miss carveout poses. It submits that it is not relying on the shortfall from projections on its own to avoid the contract but is relying on them to demonstrate the magnitude of the MAE caused by the pandemic. This brings me to the next carveout.
ii. The Emergency Carveout
98Section 1.1 (ccc) carves out from the definition of MAE any material adverse effect that results from or is caused by:
(i) worldwide, national, provincial or local conditions or circumstances, whether they are economic, political, regulatory (including any change in Law or IFRS) or otherwise, including war, armed hostilities, acts of terrorism, emergencies, crises and
natural disasters.
99Duo submits that this carveout does not apply because it does not include pandemics. In support of this argument, Duo and its experts cite studies showing that the use of a pandemic exclusion increased to 72% of agreements with crisis carveouts compared to 33% in 2019. Duo adds to this the maxim expressio unius est exclusio alterius (“the expression of one thing is the exclusion of the other”) and argues that the omission of pandemics from the list of enumerated carveouts in section 1.1 (ccc) reflects an intention to allow Duo to avoid the SPA by reason of the pandemics. I do not share this view.
100The emergency carveout is broadly worded. Its wide ambit is consistent with a general principle underlying MAE clauses of allocating systemic risks to the purchaser while leaving company-specific risks with the seller.41 That makes eminent commercial sense. The seller is more likely to have control or influence over company specific risks. A company-specific risk also, by definition, changes the nature of what the purchaser is buying. Systemic risks are different. They are ones over which the vendor has no influence or control. Systemic risks are part of the more general risk any entrepreneur assumes when running a business. If a purchaser wants to protect itself against systemic risks, it can do so by defining the MAE to include a deterioration in financial condition of the company regardless of cause. The parties did not do that.
101As noted earlier, Covid-19 was known to Duo and its investors. Even in the face of that knowledge Duo agreed to an extremely broad emergency carveout. The broad language at the beginning of the carveout clearly would include a pandemic. A pandemic is clearly a “worldwide, national, provincial or local condition or circumstance.”
102Duo’s focus on the expressio unius est exclusio alterius maxim is misplaced. After referring to “worldwide, national, provincial or local conditions or circumstances, whether they are economic, political, regulatory (including any change in Law or IFRS) or otherwise” are the words “including war, armed hostilities, acts of terrorism, emergencies, crises and natural disasters.” The plain language meaning of the word “including” is that it embraces what follows but does not do so exclusively. If there were any need to make that clear, s. 1.06 of the SPA does so by providing that:
“the words ‘includes’, ‘including’ and similar expressions mean ‘includes (or including) without limitation’”.
103The carveout specifically captures “emergencies” or “crises”. The COVID-19 pandemic has regularly been referred to as an “emergency” by federal, provincial and municipal authorities. The federal government has implemented relief programs like the Canada Emergency Wage Subsidy, the Canadian Emergency Response Benefit (“CERB”), the Canada Emergency Commercial Rent Assistance, and the Canada Emergency Business Account.42 Provincial governments have declared states of emergency, public health emergencies, and made orders under emergency legislation, including Ontario’s Emergency Management and Civil Protection Act.43 Municipal governments have implemented emergency orders and by-laws. 44
104It is by no means unreasonable for a purchaser to agree to assume the risk of a pandemic. The effect of the pandemic on Fairstone is similar to the effect of a recession. Duo knew it was buying a cyclical business that would encounter downturns from which it would recover. The risk of economic contractions is one of the general risks any business owner assumes. Assuming the risk of a pandemic is consistent with the general concept that MAE clauses protect the purchaser against company specific risks but not against systemic risks.
iii. General Market Change Carveout
105Section 1.1 (ccc) also carves out from the definition of MAE:
(ii) changes in the markets or industry in which the Acquired Companies operate;
106Duo’s expert, Daniel Fischel, states that Fairstone experienced a decline after the SPA was signed which decline was “consistent with market commentary on the Canadian consumer finance sector over the same period.” He makes further observations on the general rise in unemployment, the decline in economic activity and the “long-term negative impacts on Canadian consumers.” On Mr. Fischel’s own evidence, these are changes in the “market or industry” in which Fairstone operates.
107In closing argument, Duo’s counsel submitted that the carveout was intended to capture risks to a specific industry rather than the world at large. In Duo’s submission the carveout would, for example, capture a change in regulations applicable to a particular industry but not a general market downturn.
108That is not, however, what the carveout says. It refers simply to changes in the markets or industry in which Fairstone operates. The fact that other markets or industries may be experiencing the same changes does not in any way diminish the fact that there is a change in Fairstone’s market or industry. Indeed, throughout the trial, there was much evidence about the general tightening of lending during economic downturns. That more general change to the consumer lending market would, for example, result in decreases to Fairstone’s year-over-year loan origination numbers. To the extent that decrease is caused by a change to the consumer lending market, it does not constitute a MAE. This approach is also consistent with the more general directional concept that systemic risks generally belong to the purchaser unless otherwise specifically provided
iv. Materially Disproportionate Adverse Impact
109Fairstone has met its evidentiary burden and has established that all three carveouts apply.
110The fact that any adverse effect on Fairstone was caused by an emergency or general market changes is not, however, the end of the MAE analysis. The emergency and general market change carveouts remain outside the definition of MAE only if they
“do not have a materially disproportionate adverse impact on the Acquired Companies relative to other Persons in the industries or markets in which the Acquired Companies operate.”
111As a result, the next question becomes whether the Covid-19 pandemic or the changes in the markets or industry in which Fairstone operates have had a materially disproportionate impact on Fairstone relative to others in its industries or markets. Duo bears the burden of proof on this issue because it is the party making the assertion.
112The answer to this question turned largely on an assessment of three experts the parties tendered at trial. I review their evidence below. All three were highly qualified. Neither party contested the qualifications of an opposing expert. I prefer the evidence of Fairstone’s experts, Michael Andrews and Walter Lubiana over that of Duo’s expert, Daniel Fischel. Messrs. Andrews and Lubiana dealt with the issue of disproportionate effect head on. Mr. Fischel dealt with it indirectly. I had the sense that Mr. Fischel had devised an analytical method that would allow him to provide an answer helpful to Duo’s case as opposed to dealing directly with the question at hand.
113For the reasons set out below in my analysis of the evidence of Messrs. Andrews and Lubiana, I have concluded that, on the record before me, neither the Covid – 19 emergency nor the changes in the market or industry that have occurred since the SPA was entered into have had a materially disproportionate adverse impact on Fairstone relative to others in the industries or markets in which it operates.
a) Michael Andrews
114Michael Andrews provided three reports on behalf of Fairstone. He concluded that Fairstone was not disproportionately affected by the financial events attributable to the Covid pandemic and in many respects outperformed others in the market.
115Mr. Andrews has spent his entire career in various segments of the financial services industry. Early on in his career he worked as a personal and commercial lender with TD Bank. He then worked with the Deposit Insurance Corporation of Ontario which, at the time, was the regulator for Ontario’s credit unions. Following that he spent four years as Director of Research with the Conference Board of Canada where he managed programs in financial services and tax policy. He then served as a member of the McKay Task Force which conducted a broad review of Canada’s financial sector, an advisor to the Director of Competition and as a Financial Sector Advisor with the International Monetary Fund. In this latter capacity he provided technical assistance in financial sector policy, legislation, regulation, banking supervision and problem bank resolution. Since 2003 he has acted as a financial consultant and has had over 100 retainers from the International Monetary Fund including retainers involving the introduction of IFRS 9 which significantly changed the way in which loan-loss provisions were calculated and recorded. He is on the IMF Roster of Experts for capital markets and financial institutions.
116I found Mr. Andrews to be fair-minded and objective. He readily admitted to weaknesses or errors in his reports. He was neither argumentative nor defensive. His experience was direct and relevant.
117Mr. Andrews compared Fairstone to its most direct competitor, Goeasy Ltd., and others in the industry with reference to a variety of parameters including net income, expenses, impairment charges, operational expenses as well as the quality of the portfolio and its history of managing problems effectively. When considering those factors as a whole, Mr. Andrews formed the view that Fairstone has not been and was highly unlikely to be disproportionately affected by the pandemic. He considered those parameters as of May 27, 2020, and after.
118By way of example, Mr. Andrews compared Fairstone against comparators for net income for the first two quarters of 2020. Fairstone compares favourably to others in the market and industry, including Goeasy as figure 1 from his report of August 28, 2020 demonstrates:
119Mr. Andrews also compared Fairstone to its competitors across a broad range of qualitative and quantitative factors. That comparison demonstrated that Fairstone’s portfolios were more robust, better able to withstand market shocks and less likely to suffer in the future than Goeasy or others in the same market. His additional observations included the following:
(i) Fairstone provided fewer payment deferrals and other borrower accommodations than did Goeasy or the majority of other comparators. In April 2020, 12% of Goeasy’s customers used some form of payment accommodation compared to 7 – 8% in a typical month. Fairstone’s typical monthly request for borrower accommodations is 3.4%. In April 2020 Fairstone’s accommodations rose to 4.24%.
(ii) Fairstone historically writes of approximately 6% of its loan portfolio. Goeasy writes off approximately 13%.
(iii) Fairstone’s portfolio includes a higher proportion of homeowners than that of Goeasy.
(iv) Fairstone’s clients also had a higher average credit score than those of Goeasy.
(v) Fairstone’s portfolio contains insurance against unemployment for twelve-month period. Goeasy’s portfolio provides insurance against unemployment for only 6 months.
(vi) In May 2020, Goeasy received $7.7 million in credit insurance payments. Fairstone received $3.7 million. Given that Fairstone’s loan portfolio is approximately 2 ½ times larger than that of Goeasy but Fairstone received only one half of the credit insurance payments that Goeasy did, Fairstone’s portfolio would appear to be more robust than that of Goeasy.
120These factors constitute concrete evidence that allow me to reach an informed judgment about the proportionate effect of the pandemic on Fairstone compared to others that is tethered to realities and not to mere possibilities.
121In Mr. Andrews’ view it was relevant to consider events after May 27 because Duo’s experts had taken financial data up to May 27, including, loan-loss reserves, and had prorated that data for the balance of the year to conclude that Fairstone would be disproportionately affected by the pandemic. That approach is flawed in light of reporting requirements under IFRS 9. IFRS 9 requires a company to take reserves for losses as soon as they become aware that losses might arise. This contrasts with GAAP which required a company to take a reserve when the losses actually arose.
122In April 2020 Fairstone projected a loan loss reserve of $35 million in the second quarter. Duo’s experts assumed that a similar sized reserve would be taken in the remaining quarters of the year to demonstrate that Fairstone was disproportionately affected by the pandemic.
123This has several flaws. First it runs the risk of double counting. If the point of IFRS 9 is to project losses, then the losses taken in the first quarter are meant to anticipate losses in subsequent quarters. Stopping the analysis as of May 27 also punishes those companies, like Fairstone, that take a conservative view of loan-loss reserves and rewards companies that take an unrealistically optimistic view of loan-loss reserves. Finally, looking at actual results between May 27 and the end of August provides a good benchmark against which to test whether the assumptions of Duo’s experts were realistic.
124Second quarter actual results demonstrate that all lenders were overly conservative in their first quarter loan-loss reserves and projections for future reserves. As Mr. Andrews noted, almost all members of the comparator group to which he compared Fairstone reduced loan-loss provisions for the period ending June 30 or July 31.
125When Mr. Andrews compared Fairstone and Goeasy over a number of financial metrics for the quarter ended March 30, 2020 and quarter ended June 30, 2020, Fairstone came out ahead on all metrics except portfolio growth. Mr. Andrews paid little attention to loan origination as a comparative metric because, as he put it, originating new loans during an economic downturn is easy. All you have to do is lend money to people others will not lend to. That does not, however, bode well for long-term profitability.
126During cross-examination, Duo’s counsel took Mr. Andrews to task for the manner in which he defined the relevant market for comparison purposes. By way of example, counsel criticized Mr. Andrews for failing to follow Competition Bureau guidelines about market definition. I place no weight on that criticism. The purposes for which the Competition Bureau Guidelines define a market are different from those for which the comparison is being done here.
127Mr. Andrews readily conceded the difficulties of establishing a comparator group for Fairstone because of the absence of public companies in the specific market space that Fairstone services. While there is a good deal of information available about public companies, there is little if any information available about private companies.
128Mr. Andrews chose to compare Fairstone first against its primary competitor, Goeasy, which is public, and then looked to market participants that Goeasy had defined as its competitors.
129While the selection of any comparator group is imperfect, Mr. Andrews’s choice made sense. He continued to compare Fairstone against its primary competitor, but sought to broaden the comparison in the event a single comparator generated false results.
130Duo’s own advisors noted the difficulty of finding comparators before the litigation arose. As part of the acquisition exercise, Duo retained PWC to prepare a due diligence report. In its report, PWC noted that Fairstone was one of the largest near prime lenders and that “there are no truly comparable Canadian public lenders with a book a business similar to the Company.”
b) Walter Lubiana
131Walter Lubiana was Fairstone’s second expert. He provided what he described as a qualitative analysis of the likelihood that Fairstone would be disproportionately negatively affected relative to others in the market.
132Mr. Lubiana spent most of his career at Household Finance and at HSBC Finance after it acquired Household Finance. He joined Household Finance in 1981 and became CEO in 2010. He left HSBC after it exited the market in 2014.
133Mr. Lubiana notes that Fairstone has significant advantages over other participants in the nonprime credit market. It has successfully navigated through economic cycles for almost 100 years. It successfully weathered the 2008-2009 financial crisis and the 2015 downturn in Alberta when other market participants did not. It has a seasoned senior leadership team, a significant tenured employee base with experience in handling subprime customers and an extensive proprietary database accumulated and tested through past economic cycles.
134Mr. Lubiana compares this to other nonprime lenders who are more recent entrants to the market and whose leadership team and employees are less experienced. By way of example, Fairstone’s managers have an average tenure of 18 years. Its branch employees have an average tenure of 8 years. Mr. Lubiana notes that most of Fairstone’s competitors have not even been in the market that long.
135Although Duo’s counsel criticized Mr. Lubiana for having no analytical data to support these qualitative measures, I note that Duo and its advisors came to similar conclusions before deciding to purchase Fairstone. By way of example, Duo made a presentation to its lenders on February 24, 2020 in which it highlighted that Fairstone had:
a “proven history of through-the-cycle of success”
a “100-year operating history with demonstrated resilience through cycles.”
“Unmatched proprietary data & analytics provide competitive advantage”
“Dynamic risk management approach based on customer performance and behaviour through multiple cycles is reflective of Big Bank risk management strength.
Unique, non-commoditized and proprietary “through the cycle” data set acts as a competitive moat and barrier to entry in the market.
136These qualitative measures are particularly relevant when considering whether a MAE “would reasonably be expected” under the language of section 1.1 (ccc). Given the deep experience of Fairstone’s management in navigating through economic cycles and the value Duo itself placed on that experience it is unlikely that the pandemic will have a materially disproportionate adverse impact on Fairstone relative to others in its markets.
c) Daniel Fischel
137Mr. Fischel was Duo’s expert on the issue of disproportionate effect. He is highly experienced. Mr. Fischel has served as Dean of the Chicago Law School, Director of its Law and Economics Program and has taught at the University of Chicago Business School. He has taught corporate finance classes on principles of valuation and the effect of market conditions on industries. He has been cited by courts hundreds of times throughout the United States. He is currently the Chair and President of Compass Lexecon, a well-known global consulting firm specializing in the application of economics to legal and regulatory issues.
138Mr. Fischel concludes that Fairstone has been disproportionately adversely affected by the Covid pandemic. He reaches this conclusion by noting that Fairstone’s decline in performance as of May 27, 2020 is worse than the decline in performance of others in the industry.
139I do not find Mr. Fischel’s opinion persuasive and prefer the opinions of Mr. Andrews and Mr. Lubiana. Unlike Mr. Andrews, Mr. Fischel did not compare actual operational parameters of Fairstone against those of others in the industry. Instead, Mr. Fischel based his opinion on a comparison of Fairstone’s results as of May 27, 2020 against the median result derived from analysts’ projections for public companies for a similar period. The analysts’ projections focus almost exclusively on net operating income.
140Mr. Fischel’s approach has several limitations.
141Mr. Fischel focused on comparing analysts estimates of earnings per-share. This is a somewhat frail basis for comparison in an economic downturn. Earnings-per-share can be heavily influenced by the size of loan-loss reserves a company takes. This punishes conservative companies over unrealistically optimistic ones. In addition, it runs the potential of rewarding companies that write what are ultimately improvident loans. Increasing the size of one’s loan portfolio may have a short-term positive impact on profit. If, however, those loans are ultimately improvident in a down turning economy, they turn into heavy losses in the future.
142To continue with the concept of loan-loss reserves, in April 2020 Fairstone had prepared a new forecast at the request of Duo and OSFI which tried to take into account the effects of the pandemic on its business. That forecast envisaged Fairstone taking a loan loss reserve of $35 million in the second quarter. A reserve of that size was projected to result in a not insubstantial loss for the second quarter. Mr. Fischel then compared the projected loss for the quarter against the consensus estimates prepared by analysts for public companies in Fairstone’s industry or market. The results of those analysts’ reports are, however, influenced by the date on which the reports were prepared. If they were prepared before the company took Covid into account, they would result in lower loan-loss reserves and higher net operating income. Mr. Fischel admitted in cross-examination that Fairstone’s projected loss was driven by the loan loss reserve that it projected taking at the end of the second quarter.
143Fairstone did not, however, take the $35 million loan loss reserve at the end of the second quarter. By June, Fairstone had issued a refreshed projection which took into account the effect of significant government benefits to those affected by the pandemic. The beneficial effect of those programs reduced credit defaults considerably. Those programs put more cash into the hands of the unemployed than simple unemployment insurance did and thereby reduced the detrimental effect that unemployment historically had on the ability of the unemployed to service loans.
144Mr. Fischel did not take into account the June forecast or Fairstone’s actual second-quarter results even though that information was available when he prepared his report.
145When it came time to analyse the second-quarter results of comparators, Mr. Fischel stated that none of the comparators had reported second-quarter results by May 27, 2020 as a result of which he based his opinion on analysts’ forecasts. He did so even though actual second-quarter results were available by the time he prepared his report.
146Mr. Fischel’s results are summarized at Exhibit 5 of his first report of July 27, 2020. It shows Fairstone’s operating income having decreased by 122% year-over-year at the end of the second-quarter compared to a decrease in income by Goeasy of 4% and a median decrease of 31.4%.
147Mr. Andrew re-created Mr. Fischel’s Exhibit 5 using the actual second-quarter results that were available when Mr. Fischel prepared Exhibit 5. The actual results show Fairstone with a 43% increase in year-over-year in operating income while Goeasy’s income was down 4% and the median income in the group was down by 51.8%.
148Mr. Fischel explains that he did not take the second-quarter results into account because there was a dispute between the parties about the propriety of Fairstone not taking the $35 million reserve at the end of the second quarter. Mr. Fischel tries to cast doubt on Fairstone’s decision not to take the reserve by noting in paragraph 48 of his July 27 report that other firms in the same industry or market had either maintained or increased their expected credit losses over this period whereas Fairstone had decreased its reserves. That conclusion is, however, divorced from reality. The conclusion is based on what analysts estimated the public companies in the comparator group would take as reserves at the end of the second quarter, not on what they actually did. The estimates of loan-loss reserves that analysts used for their June end projections tended to be the same as what they estimated as of May 27. Actual results show a very different picture. The actual results were summarized in table 3 of Mr. Andrews report of August 28, 2020. That table shows that Fairstone reduced its projected loan-loss provision in the quarter ending June 30 or July 31 by 36%. The median reduction of the cohort group that Mr. Fischel analysed reduced their loan-loss reserves by 43.6% Goeasy reduced its loan-loss reserves by 49.2%. As a result, Mr. Fischel’s focus on analysts’ projections rather than on actual results shows a picture that is materially distorted from reality.
149Apart from the foregoing, I have greater confidence in the opinions of Messrs. Andrews and Lubiana because their opinions were based on analysis they did themselves. Mr. Fischel’s report appears to have been prepared by a team within his office. Mr. Fischel was assisted by three people in writing the report but those three were assisted by others that Mr. Fischel cannot identify. Mr. Fischel did not actually read the analysts reports on which his report is based and does not know who did.
150Much was made at trial about a straight comparison between Goeasy and Fairstone using certain financial parameters. Those financial parameters showed Goeasy’s metrics superior on an absolute basis to those of Fairstone. Duo suggests that this demonstrates that the pandemic has had a disproportionately adverse effect on Fairstone.
151I cannot agree with that analysis. In recent years, Goeasy’s absolute results such as portfolio growth or net income growth have been superior to Fairstone’s. Goeasy is a younger, less mature business aimed at a slightly different market which is growing rapidly. The continuation of that trend into the pandemic merely reflects the historical relationship between Fairstone and Goeasy. It does not demonstrate that Fairstone has been disproportionately adversely affected.
Conclusion on MAE
152I have concluded that Fairstone falls within the projections miss, emergency and industry change carveouts of the MAE clause and that Fairstone was not and was not reasonably expected to be disproportionately affected by the pandemic relative to others in the market or industry. As a result, there is no MAE as defined by the SPA.
153At best, what occurred here is that Duo had agreed to purchase Fairstone at what might turn out to have been the top of the market if in fact we are entering a longer economic downturn. MAEs are not generally designed to protect purchasers against the vicissitudes of market timing. Although features can be built into a MAE that protect purchasers against bad market timing, the SPA contains no such features.
III. Ordinary Course Covenant
154Section 5.1 of the SPA requires Fairstone to operate the business in the ordinary course between the date the SPA was signed and the date of closing. More specifically, it provides:
Except as expressly provided in this Agreement or the Disclosure Letter (including in Schedule 5.1 of the Disclosure Letter) or with the prior written consent of Purchaser, which consent shall not be unreasonably withheld, and to the extent lawfully able to do so, the Principal Sellers shall cause the Acquired Companies to, and the Acquired Companies shall, during the Interim Period, conduct their respective Businesses (as applicable) in the Ordinary Course and, without limiting the generality of the foregoing, the Principal Sellers shall cause each Acquired Company not to:
155This is followed by a list of 23 specific prohibitions describing steps that Fairstone may not take without Duo’s consent.
156“Ordinary Course” is defined in section 1.1 (mmm) as:
(mmm) “Ordinary Course” means with respect to an action taken by a Person, that such action is consistent with the past practices of the Person and is taken in the ordinary course of the normal day-to-day operations of the Person.
157Duo submits that Fairstone took five specific steps that breached the covenant in section 5.1 namely that Fairstone (i) made changes to its branch operations model, (ii) made changes to its collection process by offering extensive deferment programs, (iii) changed its employment policies, (iv) changed its expenditures and (v) changed its accounting methods. In addition, Duo complains of a number of other miscellaneous changes.
158In my view, Fairstone has not breached the ordinary course covenant. Fairstone’s conduct falls within the meaning of ordinary course as defined by the SPA and the case law. Moreover, if the conduct complained of did fall outside of the ordinary course of business, it is conduct to which Duo would have been required to consent had it been asked.
159I will proceed in the analysis first by examining the case law interpreting ordinary course covenants and then by reviewing the evidence surrounding the steps that Duo alleges are outside of the ordinary course of business.
A. Applicable Legal Principle
i. Level of Materiality
160Section 5.1 is a covenant. It is turned into a closing condition by s. 6.2. Section 6.2 (a) (iii) requires that all covenants be performed by the seller “in all material respects.” As a result, it is not necessary for Fairstone to operate the business in the ordinary course without any deviation whatsoever. It is only necessary for Fairstone to have operated the business in the ordinary course “in all material respects”.
161In Akorn, the Delaware Court of Chancery held that the reference to “all material respects” did not incorporate the common law doctrine of material breach. That standard would give the purchaser an exit only if the breach were so material that it went to the root or essence of the contract. In the context of an acquisition, the reference to “in all material respects” is meant to indicate that a minor deviation from ordinary course business will not excuse the buyer from closing. In Akorn the court held that the reference to all material respects was meant to capture breaches that are significant in the context of the parties’ contract even if they would not rise to the common law standard of going to the root of the agreement. Rather, the materiality requirement is aimed at determining whether the change in question would be viewed by a reasonable purchaser as having altered the “total mix” of information so as to lead a reasonable purchaser not to acquire the business or to acquire it on significantly different terms.45
162I adopt that approach here. It is consistent with the contextual interpretation of the ordinary course covenant (discussed below) and is consistent with the principle of interpreting the covenant from the buyer’s perspective.
163Given that I have found that Fairstone did not operate outside of the ordinary course, the materiality requirement does not create any particular issues in this case. If I am wrong in this regard and Fairstone did operate outside of the ordinary course, I do not find that any of its conduct breaches the ordinary course covenants in a material respect because, with the exception of alleged changes to accounting methods, the conduct was temporary, had been terminated by any closing date, had no long-lasting effect on the business and imposed no obligations on Duo. If I am wrong about the alleged changes to accounting methods and, in fact, those alleged changes amounted to operating outside of the ordinary course, then in my view that was also not material because Duo was fully protected against the alleged changes at issue here by section 2.7 of the SPA.46
ii. Definition of Ordinary Course in Case Law
164The expression “ordinary course” arises in different legal contexts including insolvency legislation, improper preferences, and business agreements. The context in which the phrase arises may influence the meaning. It may nevertheless be helpful to consider factors that courts have taken into account in contexts different from the case at bar because those factors may help shed light on what circumstances may be relevant to the case at hand.47
165In Re Pacific Mobile Corp.48, the Supreme Court of Canada held at para. 3:
It is not wise to attempt to give a comprehensive definition of the term "ordinary course of business" for all transactions. Rather, it is best to consider the circumstances of each case and to take into account the type of business carried on between the debtor and creditor.
166Whether a transaction is in the ordinary course of business is to be determined from the buyer’s perspective.49 As a noted American text puts it:
The Authors believe that the Seller’s concerns here when taken in the context of a business it has agreed to sell generally are not as important as the Buyer’s. Thus, the equities of the situation generally weigh in on the side of the Buyer.50
167The purpose of ordinary course covenants is to ensure that the business the buyer pays for at closing is essentially the same as the one it decided to buy when signing the agreement of purchase and sale. 51 In addition, ordinary course covenants aim to mitigate or eliminate the moral hazard of sellers acting in their own interest to the detriment of the purchaser between signing and closing.52
168The idea of ensuring that the business at closing is essentially the same as the one the purchaser decided to buy when signing the agreement requires some nuance. Conduct that renders the business different in an essential way at closing than it was at the time of signing falls afoul of an ordinary course covenant. However, as the Ontario Court of Appeal pointed out in, Stelco, conduct that falls short of changing the business in an essential way can nevertheless constitute a breach of an ordinary course covenant.53 Thus, in Stelco, the Court of Appeal found that a transaction that required Stelco to pay $320 million over 10 years for IT outsourcing was out of the ordinary course even though it did not change the essential nature of the business because it imposed a material, long-term and far-reaching obligation on the business.
169In coming to this conclusion, the Court of Appeal relied on the dictionary meaning of “ordinary” as being “of common occurrence, frequent, customary, usual” and “of the usual kind, not singular or exceptional” and the definition of “course” as meaning “habitual or ordinary manner of procedure; way, custom, or practice.”54 The court then listed several non-exclusive factors to consider when determining whether conduct is outside of the ordinary course:
a) Is the conduct distinguishable from the normal course of the company’s operations because of its particular complexity or its far-reaching or otherwise unusual nature?
b) Did it arise out of some special or peculiar situation?
c) Did it require approval from the company’s shareholders or board of directors?
d) Did it require special notice by the company to anyone?
e) Was the conduct unusual or isolated as opposed to routine?
f) Does the conduct reflect standard practice in the industry?55
170In Bankruptcy and Insolvency Law of Canada, the learned authors note that, in the insolvency context at least, the analysis should ask whether ordinary businesspersons would have acted in the same way or whether they would consider the activity anomalous or unbusinesslike:
The approach of the court should be objective, i.e., the court must ask itself whether, regarded objectively, solvent persons would, in the normal course of business, have acted in the same way as the parties concerned in the impugned transaction. In other words, is the transaction one that would not to the ordinary businessperson appear anomalous or unbusinesslike or surprising?56
171In 369413 Alberta Ltd. v. Pocklington,57 the Alberta Court of Appeal said the following about ordinary course covenants:
Courts are required to undertake a broad and case-specific analysis of the ordinary course of a particular company’s business, for it is considered unwise to attempt a comprehensive definition. The analysis is to be achieved through an objective examination of the usual type of activity in which the business is engaged, followed by a comparison of that general activity to the specific activity in question. The transaction “must fall into place as part of the undistinguished common flow of business carried on, calling for no remark and arising out of no special or peculiar situation.
In order to determine whether a transaction was in the ordinary course of a company’s business, a court must consider all the circumstances which were known, or ought reasonably to have been known, by the parties at the time. Courts have identified a number of factors which may be taken into account:
(i) The nature and significance of the transaction: it ought to be one that a manager might reasonably be expected to carry out on the manager’s own initiative without making prior reference back or subsequent report to superior authorities, such as the board of directors or the shareholders;
(ii) The value of any asset sold: the disposition should have been made with proper regard to its value;
(iii) The quantity of assets sold: the transaction ought not to resemble a liquidation of assets;
(iv) The reason for the transaction: it ought not to have occurred as a response to financial difficulties or in suspicious circumstances; and
(v) The intent of the transaction: neither its intent nor its effect should have been to undermine bank security.
To this list I would add:
(vi) The frequency of the type of transaction: an unusual or isolated transaction might be viewed differently from a routine one; and
(vii) The arm’s length nature of the transaction: a transaction between a company and a party with whom it is related should receive careful scrutiny.58 (Citations omitted)
172In Canadian Commercial Bank v. Prudential Steel Ltd.,59 Virtue J. of the Alberta Court of Queen’s Bench held that the requirement to contextualize ordinary course meant that one can consider what is ordinary in light of the circumstances, including the difficult financial circumstances that a business may find itself in at the time of the impugned transaction:
In determining what is “in the ordinary course of trade”, I am of the view that one can consider what is ordinary in the light of all the circumstances, including the difficult financial condition in which a business finds itself at the time of the transaction impugned.
173If an objective viewer would reasonably conclude that the impugned transaction is one which a company would ordinarily do in the course of trade, including a transaction made to keep its business going in difficult times, then so long as the transaction is a bona fide one made for the purpose of carrying on the business and not one which either the vendor or the purchaser intended to be primarily for the purpose of preferring one creditor over another, it meets the test of s. 6. It would, of course, still remain for the creditor to show that “the money paid or the goods or other property sold or delivered bear a fair and reasonable relative value to the consideration therefor.’’60
174This passage arguably conflicts with the passage cited from the Alberta Court of Appeal in Pocklington above. The court in Prudential Steel clearly thought it was appropriate to assess the ordinary course nature of the conduct in light of the overall financial context in which the business found itself. In Pocklington, however the Alberta Court of Appeal stated that the conduct “ought not to have occurred as a response to financial difficulties…” This raises the question of how the court should proceed in assessing whether a company’s conduct falls within the ordinary course when that conduct is undertaken in response to financial difficulties.
175The issue is of particular relevance here because Duo argues that certain steps Fairstone took in response to Covid - 19 were outside of the ordinary course because they were taken in response to financial difficulties and were unusual in the sense that they differed from Fairstone’s conduct before signing.
176How then to resolve that issue? Without purporting to set out a universal rule, it may be helpful to set out certain principles.
177As a general rule, the purchaser of a business accepts systemic economic risks associated with the ownership of a business. This, for example, includes the risk of economic contractions and the detrimental effect they have on a business. In the absence of specific language in an agreement that protects a purchaser against economic slowdowns, a purchaser would generally be seen to accept the risk of those slowdowns between the time of signing and the time of closing. As noted, when discussing MAEs earlier on, there are many ways that purchasers can protect themselves against systemic economic risks between signing and closing.
178It is part of the ordinary course of any business to encounter local or national recessions and to take steps in response to those sorts of systemic economic changes. When such systemic changes occur, the concept of ordinary course is more faithfully interpreted by comparing what the business has done in similar economic circumstances to what it is doing now or by comparing what the business is doing now to what other businesses are doing. The analysis must also be alert to the magnitude and duration of the changes. By way of example, selling an entire division of a business would probably be outside of the ordinary course even if the sale were motivated by a good faith effort to combat a serious recession.
179If, however, the business takes prudent steps in response to an economic contraction, that have no long-lasting effects and do not impose any obligations on the purchaser, it should not be seen to be operating outside of the ordinary course. To hold otherwise risks turning the ordinary course covenant into a guarantee of local or national economic performance over which the seller has no control. While buyers can obtain such protection, it is preferable that it be obtained through specifically negotiated clauses by more precise references to identified financial parameters rather than through a simple requirement to operate in the ordinary course.
180In this light, a seller’s response to financial difficulties may remain in the ordinary course if they respond to broader systemic economic contractions61 and do not fall afoul of other principles relevant to the interpretation of “ordinary course.” The more specific the economic difficulties are to the seller, however, the less likely the conduct is to fall within the ordinary course.62
181This interpretation achieves the fundamental purpose of the ordinary course covenant which is to protect the purchaser against company specific risks and the moral hazard of management acting in a self-interested, opportunistic manner detrimental to the purchaser’s interests.
182I glean the following principles from the foregoing summary of the case law. There is no single definition of an ordinary course obligation even in the context of a share purchase agreement like the one at hand here. The specific interpretation of an ordinary course obligation will depend on the circumstances of the case at hand. A proper analysis takes into account and balances all of the relevant circumstances some of which may include the following:
Does the conduct render the nature of the business different at closing than it was at the time of signing?63
Does the conduct give rise to moral hazard concerns?
Was the conduct arm’s length in nature?
Is the conduct part of the usual, habitual flow of the business or was it unusual?
If the conduct was unusual, was it ordinary in light of the circumstances the business was facing?
Were those circumstances systemic or specific to the company?
How does the conduct compare to standards in the industry?
What was the intent behind or reason for the conduct?
Was the conduct pursued in good faith for the purpose of continuing the business?
Does the conduct defeat the legitimate interests of a creditor or other interested party?
Would the conduct surprise a reasonable businessperson?
What was the magnitude and duration of the conduct? Does it have a long-term impact?
Are there equities that should weigh in favour of the purchaser?
183Not all of these factors will be relevant in any particular case. In some cases certain factors will assume greater importance than in others.
184In considering these factors courts should be mindful of an ever present underlying tension. On the one hand, courts must guard vigilantly against moral hazard and opportunistic behaviour by sellers. At the same time, they must be mindful of similar opportunism on the part of purchasers. Ordinary course covenants are generally not intended to protect purchasers against the risk of market timing unless they contain specific language to that effect.
B. Application of These Principles to the Clause at Issue
185Duo submits that most of the factors listed above are irrelevant here because the definition of ordinary course in section 1.1 (mmm) of the SPA requires that the conduct be “consistent with the past practices” of Fairstone AND that it be “taken in the ordinary course of normal day-to-day operations.”
186As such, Duo argues if the conduct at issue differs from Fairstone’s practice before the contract was signed, it is outside of the ordinary course. Duo furthermore argues that any changes Fairstone made in response to the pandemic were not taken in the ordinary course but were taken in response to an extraordinary situation. This submission gives rise to three questions:
i. How should the ordinary course and the MAE covenants be read together?
ii. What is the meaning of “consistent with past practices”?
iii. How do the principles articulated in the previous section apply to the facts at issue here?
i. The Ordinary Course and the MAE Covenants
187In effect, Duo argues that nothing done in response to the pandemic is ordinary course because the pandemic is an extraordinary event. Fairstone submits that Duo’s position negates s.1.1(ccc) (i) which provides that any adverse effect caused by an emergency does not amount to a MAE and does not give Duo the right to avoid the contract.
188Duo responds by pointing out that it should be afforded “the protections for which they bargained.”64 If Duo bargained for some protections in the MAE covenant and others in the ordinary course covenant, it should be entitled to whatever protection the language of either covenant affords. While I agree with this proposition in principle, it must be tempered by the competing rule of contractual interpretation that contracts should be read as a whole.65
189Duo’s position effectively means that Fairstone could not operate in the ordinary course during the pandemic because anything it did to respond to the pandemic would be operating outside of the ordinary course. This would in effect, make the pandemic a reason for not closing the transaction even though emergencies in the nature of the pandemic are excluded from the definition of MAE.
190Given the specificity of the emergency exclusion in the MAE clause, I do not think it would be appropriate to use the more general ordinary course provision to, in effect, override the more specific MAE provision. To do so would not read the contract as a whole but would read it as a series of unrelated, standalone provisions.
ii. What is the meaning of “consistent with past practices”?
191Duo correctly notes that to fall within the definition of ordinary course in s. 1.1 (mmm) of the SPA, Fairstone’s conduct must not only be in the ordinary course in the sense that it be taken in the course of normal day-to-day operations but also that it be “consistent with the past practices” of Fairstone.
192Duo submits that this requires me to compare Fairstone’s conduct before and after signing. Any difference, argues Duo, cannot be in the ordinary course because it is not consistent with past practice.
193It is important, however, to note that the definition of ordinary course in the SPA requires that Fairstone’s conduct be “consistent” with past practices, not that it be identical to them.
194The Oxford English Dictionary defines the usual and current sense of “consistent” as meaning:
Agreeing or according in substance or form, congruous, compatible.
Constantly adhering to the same principles of thought or action.
195Fairstone’s response to the pandemic was consistent with past practices in the sense that during past economic contractions, Fairstone took steps to reduce expenditures and tighten lending requirements. In this regard, Fairstone’s past and current practices are congruous, compatible and adhere to the same principles of thought and action.
196All changes Fairstone made were also consistent with past practices in that they allowed Fairstone to continue its normal day-to-day business operations of lending to consumers, managing the consumer loan portfolio and collecting on its consumer loan portfolio.
iii. Application of Ordinary Course Principles to the Facts
197Fairstone’s response to the pandemic does not give rise to any concerns when evaluated in light of the factors set out in paragraph 182 above.
198The nature of Fairstone’s business remained the same at closing as it did at the time of signing. None of the changes led to any fundamental modification of Fairstone’s business. In a similar vein, none of the changes involved the assumption of longer-term obligations with which the purchaser would now be saddled such as the IT contracts in Stelco. None of the changes imposed any structural modifications on the business that would be difficult to alter or that would have long term effects.
199While some of the measures Fairstone took involved doing business differently than it did before the pandemic, there is no evidence to suggest that Fairstone was acting differently than other participants in the industry. The reason for the changes Fairstone made was to keep the business running as normally as possible in light of a worldwide pandemic. The conduct was pursued in good faith for the purpose of continuing the business, not changing it. I do not think it could be fairly said that the changes Fairstone made would have surprised any objective businessperson. All changes were consistent with what other businesses were doing across Canada.
200The changes were in response to systemic challenges that the pandemic posed for the entire economy. The changes were not in response to challenges unique to Fairstone.
201Apart from the alleged changes to accounting methodology, none of the alterations Fairstone made to its business involved non-arm’s length transactions, potential moral hazard, equities that weighed in favour of Duo, created long-term obligations or created structural changes to Fairstone’s business.66 To the extent that any of the changes even remained in place at the time of closing, Duo could easily reverse them if it wanted to.
202I am strengthened in these views by the evidence of Fairstone’s expert, Walter Lubiana. Mr. Lubiana testified about what steps consumer finance businesses take in the ordinary course to manage economic downturns. He expressed the opinion that the steps Fairstone took were consistent with what one would expect a consumer finance company to do in the present circumstances and were consistent with steps he has seen other consumer finance companies take during other economic downturns. Fairstone’s measures therefore compare favourably with standards in the industry.
203Duo criticizes Mr. Lubiana because he did not compare Fairstone’s pre-Covid practices with its post-Covid practices and because he retired in approximately 2014. Duo did not, however, introduce any evidence to suggest that the steps that Fairstone took in response to Covid were any different than it would expect other players in the consumer finance industry to take. In addition, Duo introduced no evidence to suggest that Mr. Lubiana’s evidence was somehow outdated by virtue of the fact that he retired in 2014.
204The claim that potentially comes closest to creating moral hazard is the alleged change to Fairstone’s accounting methodology. The steps Fairstone took in this regard could lead the purchaser to pay a higher price than it would otherwise have to. It will be particularly important to examine those changes carefully. I have done so and have come to the conclusion that the accounting was appropriate. This issue is discussed in further detail beginning in paragraph 249 below.
205In my view the ordinary course covenant, like the MAE covenant, should be read in the context of the entire transaction. Here, Duo was acquiring a business for the long term. It was consciously acquiring a management team that was experienced in navigating through economic downturns. The fact that management put into place modest measures to navigate through an economic downturn that arose after the signing the SPA, does not amount to operating outside of the ordinary course in the context of a long-term acquisition. Long-term acquisitions will inevitably encounter economic downturns. Responding to an economic downturn is not operating outside of the ordinary course but amounts to operating with in the ordinary course of a business. One ordinary course aspect of managing a business is to respond to economic downturns.
206To take this view does not mean that businesses are now free to respond to economic downturns in a manner that would fundamentally change the nature of the business the purchaser is acquiring, impose long-term obligations on a purchaser, dispose of assets or dispose of portions of the business. Here, however, the changes that Fairstone implemented do not come anywhere near to this sort of conduct.
C. Communications with Duo
207One further benchmark against which to consider the ordinary course covenant is the communication between the parties because they may shed light on what the parties expected of each other.
208On March 4, 2020 Miriam Tawil, a director of Duo Bank and representative of Centerbridge (one of Duo’s financiers for the transaction) sent Scott Wood, President and Chief Executive Officer of Fairstone a memo that Centerbridge had circulated internally. The memo set out steps Centerbridge had asked its own staff to take and that Centerbridge wanted the companies in which it had invested to take in response to the pandemic. Ms. Tawil sent the memo in an effort to “compare notes” with Fairstone about appropriate pandemic responses. The memo is instructive. It clearly indicates that Centerbridge expected its businesses to take a number of steps in response to the pandemic. Given that Ms. Tawil was sending the memo to Fairstone’s CEO, I infer that Duo shared the views expressed in the memo and expected Fairstone to implement similar steps.
209The memo instructed Centerbridge businesses to:
Prepare a Covid – 19 business continuity and response plan
Assess and develop employee protection measures
Assess and develop customer care measures
Continually revise these plans as required
Develop activation protocols for all phases of response
Develop a cost approval process for expenses related to Covid – 19
Set up emergency remote work protocols for call centre operations, collections etc.
In addition, the memo noted:
Employees have been instructed to follow the most conservative guidelines available from the Centers for Disease Control and Prevention, World Health Organization, European Centre for Disease Prevention and Control, and other leading global and local health authorities as applicable.
210On March 10, Ms. Tawil and Mr. Wood spoke about: Fairstone’s plans in response to Covid – 19, the loss mitigation tools Fairstone employs to manage credit in financial downturns and how Fairstone handles things such as temporary branch closures during emergencies.
211On March 20 Mr. Wood wrote to Ms. Tawil referring to steps other employers were taking, including branch closures. He also set out a number of steps Fairstone proposed to implement in response to the pandemic to stay in line with the market. At the end of the email he invited Ms. Tawil to let him know if she had any concerns with Fairstone implementing the measures he proposed. Ms. Tawil responded by thanking Mr. Wood for keeping them updated. Ms. Tawil has not given evidence in this proceeding and has not suggested that others in the purchase group disagreed with the steps Fairstone proposed.
212On March 27, Mr. Wood gave a presentation to investors which included a list of steps Fairstone had taken in response to Covid – 19 and which had been discussed with Ms. Tawil. The list included the steps in respect of which Duo complains in this proceeding. Steven Smith, another director of Duo who attended the March 27 meeting said that he thought Fairstone should probably provide written notice under the SPA for some of the measures that Fairstone had implemented in response to the pandemic.
213In response to the request, Fairstone sent Duo a letter on March 27 containing four steps under the heading “Extraordinary Employment Measures due to Covid – 19”. The steps were relatively modest in nature. They provide employees with five extra personal days, waive the 90 day waiting period for current and new employees to access personal days, prolong the carryover of 2019 vacation days into 2020 and provide short-term disability coverage without a waiting period for those diagnosed with Covid. This final measure was at no cost to Fairstone because its short-term disability insurer had agreed to waive the waiting period. The additional personal days and the extended vacation carry over were implemented because Fairstone was asking employees to use personal days or vacation time to deal with any pandemic related issues that prevented them from working, such as added child care responsibilities due to school closures.
214Duo responded on April 1, 2020. The response differed in tone from earlier communications and in many ways contradicted them. Duo now took the position that Fairstone was not conducting business in the ordinary course, that Fairstone had, will or would reasonably be expected to suffer a MAE, and that Fairstone would reasonably be expected to encounter an Amortization Event. Duo took the position that the limited employment provisions referred to in the previous paragraph were implemented without seeking prior consent even though Ms. Tawil had been advised of the measures, had been invited to express concerns and expressed none. Duo then advised that:
We have not consented to any matters that are not in the Ordinary Course and believe that Sellers [ ] are in breach of their covenant in section 5.1 of the SPA to cause the Acquired Companies to conduct their business in the Ordinary Course.
We are not in a position to entertain a partial consent request with partial information. We ask that you provide us with (i) the details of all the actions that you have taken that are not in the Ordinary Course, (ii) your rationale for such actions as opposed to the alternatives and (iii) the immediate and expected impact of the pandemic and related actions taken by the business, including operationally and financially.
215This response was disproportionate to the relatively innocuous employment measures Fairstone had implemented. The only thing Duo pointed to at trial to support the argument that these measures were out of the ordinary course was that the measures differed from Fairstone’s pre-pandemic practice and that Fairstone had referred to them as “Extraordinary Employment Measures” in its letter of March 27, 2020. With the benefit of hindsight, Mr. Wood probably regrets describing the measures as “extraordinary” in his letter. However, the use of the word in this context provides more theatre than substantive legal argument to avoid a transaction worth more than $1 billion. Duo had learned of these measures earlier and had not objected. In its April 1 letter Duo did not raise any specific concerns about the measures. At trial, Duo did not explain how these particular measures changed Fairstone’s business, imposed any material financial obligation on the purchaser or how those measures were otherwise out of the ordinary course. Indeed, the focus at trial was on entirely different steps that Fairstone took in response to the pandemic.
216The letter was clearly setting the stage for Duo’s attempt to abandon the agreement. Telling Fairstone that it has already breached the SPA and then demanding details of all actions Fairstone had taken that were not in the ordinary course is to set Fairstone up for failure. This was not a good faith effort to examine whether Fairstone could manage its way through the pandemic without taking steps that were outside of the ordinary course. It was a transparent trap into which Duo hoped Fairstone might fall.
217Duo’s refusal to entertain a “partial consent request” contradicts Centerbridge’s own memo which instructed its businesses to “continually revise” its pandemic plans as required. Continual revision implicitly includes partial consents. The steps Centerbridge required its own businesses to take went considerably further than the four modest measures Fairstone proposed in its March 27 letter.
218Not surprisingly, matters deteriorated quickly. There followed a flurry of letters about the steps Fairstone was taking. Duo objected to the steps Fairstone was taking. Fairstone in turn responded that Duo had in fact implemented many of the same steps. Duo has not denied that suggestion.
D. Specific Changes to Fairstone’s Business Model
219I turn now to the specific conduct about which Duo complained at trial as being outside of the ordinary course.
i. Changes to Branch Operations and Loan Origination
220Duo submits that Fairstone changed its operation from a branch based system to an online system in response to the pandemic. According to Duo, this marked a fundamental change in the business because Fairstone distinguished itself from its competitors by the level of its branch operations. Duo described Fairstone’s branch operations as its “secret sauce.”
221Duo submits that Fairstone’s managed branches in shopping centres were closed altogether while other branches were closed as a result of suspected or confirmed Covid cases. These closures were longer than branch closures had been in the past. Remaining branches were open by appointment only with doors locked and customers able to enter only if they had arranged an appointment in advance. Appointments could only be obtained if a customer answered a number of screening questions. If a branch were closed, an alternative branch was sometimes hundreds of kilometres away. Duo also complains that Fairstone created a new digital loan processing support facility which enabled loans to be originated, processed and closed online. Before the pandemic, online loan completion amounted to less than 1% of Fairstone’s business and only existed in pilot form.
222Duo submits that changes to branch operations detrimentally affected Fairstone’s business. Duo notes that new loan origination was down 30% in March, 66% in April and 55% in May.
223Duo argues that it had agreed to buy a branch based business and was now getting an online business.
224I am unable to agree with Duo’s characterization of these changes for several reasons.
225First, Fairstone’s business did not change from a branch based system to an online system. Fairstone received approximately the same percentage of lending applications online during the pandemic as it did in the months preceding the pandemic. Branch employees continued to work in the branches. Only non-branch employees began working remotely. Of Fairstone’s 242 branches, only 2 branches and 10 smaller managed offices were closed. The balance were open by appointment between March 15 and June 14. By July 15, 2020 all but three of Fairstone’s branches were allowing customers to enter without appointments or pre-screening.
226Second, the ordinary course covenant is qualified by requiring Fairstone to conduct business in the ordinary course “to the extent lawfully able to do so.” After March 15, 2020 most of Canada went into lockdown. Businesses premises were either closed or subject to restricted entry. Although Fairstone had been categorized as an essential business and was therefore permitted to keep branches open, it was nevertheless required to operate in accordance with the recommendations and instructions of public health authorities. That included controlling access to premises so as to maintain social distancing. Other essential businesses also controlled access to premises. Fairstone had an obligation to keep its employees and customers safe. Operating exactly as it had before the pandemic would not have done so and would have exposed Fairstone to the risk of legal liability. Locking doors and requiring appointments was an easy and efficient way to maintain social distancing. The changes were of short duration and there is no evidence that they caused long-term impact on the business.
227Third, even the limited, temporary branch closures that occurred are contemplated in Fairstone’s Business Continuity Plan. Recall that Centerbridge encouraged Fairstone to develop a business continuity plan to the extent it did not have one already which plan should contemplate the closure of premises and a switch to online operations. Recall also that the memo Centerbridge sent to its own businesses stated:
Employees have been instructed to follow the most conservative guidelines available from the Centers for Disease Control and Prevention, World Health Organization, European Centre for Disease Prevention and Control, and other leading global and local health authorities as applicable.
228Allowing uncontrolled, unlimited access to branches and having all employees work in offices when remote work was possible would hardly be following the most conservative guidelines from health authorities. It would be ignoring those guidelines.
229Fourth, Duo has not demonstrated that the decline in Fairstone’s loan origination was attributable to the shift to branch appointments or online/ telephone operations. The evidence is to the contrary. Fairstone acknowledged that it tightened lending requirements and reduced exposure to risky or new loans in response to the pandemic as it has always done in response to economic uncertainty. Consumers may also have been deferring expenditures and borrowing in response to the economic uncertainty, at least in the pandemic’s early stages.
230The fact that Duo’s total credit portfolio in Canada dropped 19% between January and June 2020, when Duo has no branches and operates entirely online, supports this view.
231Finally, Duo did not introduce any evidence to suggest that the steps Fairstone took were unnecessary, inappropriate or that Duo could have improved upon them. On my view of the evidence, this was never about Duo’s disappointment at being deprived of the opportunity to co-manage any business refinements in response to the pandemic, it was about Duo trying to find a way to avoid the SPA or renegotiate it.
ii. Changes to Collection Process
232Duo submits that Fairstone changed its collection practices by implementing a nationwide program that allowed customers to defer payments. Duo notes that Fairstone had not implemented a national deferment program since the financial crisis of 2008 and has never deferred payments merely because customer income had been affected. Duo says the pervasiveness of the deferment program affected the overall quality of Fairstone’s loan book.
233The program in question allowed a customer to defer payments for up to two months as opposed to Fairstone’s ordinary discretionary deferment of one month. Payment deferrals are part of Fairstone’s ordinary course operations. They are not unusual in the near prime and subprime business in which Fairstone operates. Near and sub-prime lenders distinguish themselves from banks by their greater flexibility. Borrowers in this market segment do not have the same ability to make regular payments as bank customers do. Flexibility is an important tool that ensures repayment and develops customer loyalty. Enforcing a loan after a missed payment often leads to a significant shortfall because near and sub-prime borrowers often do not have readily exigible assets and because enforcement costs are high in relation to the loan amount.
234Fairstone’s deferment program lasted for three months. It led Fairstone to give approximately 17% of its customers payment deferrals. A historical analysis shows a one-month spike in deferrals in March 2020 followed by a rapid return to normal deferral rates in April. The number of deferrals in April 2020 was approximately equal to that of December 2019. Moreover, Fairstone’s loan deferrals in 2020 were lower than they were during the financial crisis between 2009 and 2011. In the foregoing circumstances, the deferral program does not amount to operating outside of the ordinary course of business.
iii. Employment Measures
235As part of the ordinary course covenant, ss. 5.1 (e) and (f) of the SPA specifically prohibit Fairstone from increasing the compensation or benefits payable to any employee or paying any severance or termination to any employee other than in the ordinary course.
236Duo complains that Fairstone decreased staffing by 40 employees between April and May, 2020, increased entitlement to paid personal days, removed the normal 90 day waiting period for new employees to access paid personal days, offered a daily pay premium and relaxed short-term disability eligibility all in violation of sections 5.1 (e) and (f) of the SPA.
237In my view those steps were permissible and were taken in the ordinary course.
238Fairstone reduced its headcount by 40 of 1400 employees. Mr. Wood says he assumes the 40 departures were the result of natural attrition because Fairstone did not engage in any layoffs or terminations in response to the pandemic although he had no specific knowledge of the 40 cases. Even if I assume that all 40 were the result of terminations, dismissing 40 out of 1,400 employees during a time of economic contraction does not, in my view, rise to the level of operating outside of the ordinary course. There was no suggestion that any of the employees were given severance packages that were unduly large (or any at all for that matter) or that this was in any way an effort to give the departing employees some sort of sweetheart deal before a new purchaser took over.
239The daily pay premium amounted to $50 per day paid between April 1 to May 15 for employees who did not work remotely. It aimed to compensate them for increased costs associated with personal care or transportation to work. The cost of the entire program was $1.1 million funded entirely out of Fairstone’s existing budget. It did not result in any increase to budgeted operating expenses, did not cost Duo anything and had no effect on Duo because the program had terminated by May 15, well before any closing date. The idea of a pay premium of $50 per day was also consistent with what other businesses were paying employees for not working remotely.
240I have addressed the employment measures from the perspective of the ordinary course conduct to demonstrate that they did, in my view, fall within the conduct permitted by ss. 5.1 (e) and (f) of the SPA. In addition, however, given that certain employment measures were specifically prohibited (unless in the ordinary course), I underscore that the measures taken were ones for which, in my view, it would have been unreasonable for Duo to deny consent because they were modest in nature, consistent with the market and imposed no long term obligation on the business.
iv. Changes to Expenditures
241Duo complains that Fairstone’s expenditures decreased by 10% relative to the year-to-date budget including expenditures for marketing, technology and communications. According to Duo, these expenditures help maintain asset quality and enable growth.
242For a business to decrease expenditures in response to economic contraction amounts to ordinary course conduct. Mr. Wood indicated that Fairstone made a similar reduction in expenditures during the 2008 financial crisis. The reduction in 2020 was therefore consistent with Fairstone’s past practice.
243To maintain pre-existing expenditures in a changing market environment may well be silly. By way of example, if Fairstone had budgeted for marketing expenditures aimed at subprime borrowers, incurring those costs when Fairstone was tightening its lending to subprime borrowers in response to the pandemic would make no sense. I am not suggesting that this is what actually happened, but merely giving a conceptual example of why a 10% decrease in expenditures is not outside of the ordinary course. Numerous witnesses and Duo’s own materials noted that one of Fairstone’s strengths is its ability to manage through economic uncertainty and contractions. Doing so requires the ability, among other things, to adjust budgets and business plans quickly to ensure that they remain appropriate for the changing circumstances the business faces.
244There was no suggestion at trial that the decrease in expenditures had a detrimental impact on the business, that Duo could not make those expenditures after closing if it wanted or that deferring the expenditure until after closing would have a detrimental impact on the business Duo was acquiring.
v. Miscellaneous Changes
245Duo complains about of a number of miscellaneous changes to Fairstone’s business practices including tightening customer credit requirements, limiting real estate appraisals to exterior views only, ceasing automobile auction services, and implementing a foreclosure moratorium.
246None of these fall afoul of the ordinary course covenant.
247Fairstone regularly tightens and loosens customer credit requirements in response to prevailing economic conditions. As already noted, one reason Duo was attracted to Fairstone was because it had an experienced management team that was able to adjust its operations effectively in response to changing economic conditions.
248Limiting real estate appraisals to exterior views was not a step that Fairstone imposed but one that real estate evaluators imposed. Similarly, Fairstone did not stop automobile auctions. Auctions were considered to be nonessential businesses and could not lawfully proceed. Finally, Fairstone did not impose a moratorium on mortgage enforcement; courts did because they declined to hear mortgage enforcement proceedings during the first few months of the pandemic. Once courts began accepting them, Fairstone resumed foreclosure filings. As noted earlier, the ordinary course covenants require Fairstone to operate within the ordinary course only insofar as Fairstone is” lawfully able to so.”
vi. Changes to Accounting Practices
249Duo alleges that Fairstone changed its accounting practices in violation of the ordinary course covenant. I will address that issue by describing the complaint, describing the accounting steps that occurred and analysing whether those accounting steps breached the ordinary course covenant by reference to the evidence of witnesses and experts.
- The Accounting Complaint
250As part of the ordinary course covenant, s. 5.1 (p) of the SPA specifically requires that Fairstone not:
make any change to the accounting methods, principles, classifications or practices currently used by the Acquired Companies, except as may be required by IFRS or applicable Laws or in the Ordinary Course;
251Duo submits that Fairstone changed its accounting methods in Q2 2020 by changing the way in which it calculated its loan-loss reserves during the second quarter.
252The loan loss reserve or LLR is an estimate of the losses that Fairstone expects to incur on loans going forward. Fairstone calculates its loan loss reserve each quarter. The LLR is recorded as a liability on the balance sheet thereby decreasing the overall value of the assets of the company. The LLR is also recorded on the company’s income statement. An increase in the LLR in any one quarter decreases net income for that quarter. Similarly, a decrease in the LLR in a given quarter would increase the value of the company’s assets on the balance sheet and increase the income for Fairstone in the given quarter.
253A change in the LLR has a dollar for dollar effect on the purchase price that Duo would be required to pay for the business. Duo submits that the changes Fairstone made to its accounting methods meant that Duo would have been required to pay $35 million more for the business than if Fairstone had not made the changes. As a result, any change to accounting methods gives rise to the moral hazard concerns that ordinary course covenants are designed to protect against.
254Fairstone bases its LLR on what the parties have referred to as the Base Model. One input into the Base Model is a projected unemployment rate. Projected unemployment rates factor into LLRs because higher unemployment has historically been associated with higher default rates on loans. Historically, Fairstone has used projected unemployment data from Moody’s Investor Services to feed into the Base Model. Once unemployment data and other inputs have been entered into the Base Model, it generates a projected loss figure for loans in the portfolio that are expected to go into default in the future.
255Duo alleges that in Q2 Fairstone made the following changes to its accounting methods:
i. Instead of using Moody’s unemployment projections, it reduced the month over month change in Moody’s unemployment projections by 35%.
ii. Management made further adjustments to dampen the effect of the unemployment numbers on projected loan losses between June and December 2020.
iii. In Q2, the Base LLR Model indicated that an increase of $35 million in the LLR was called for. Fairstone did not increase its LLR at all.
256In addition to the foregoing, Duo submits that Fairstone unfairly produced only one witness: its CEO, Scott Wood. Duo submits that Mr. Wood did not have adequate knowledge to testify to the use of the accounting models Fairstone used. I do not think that is a fair characterization. It should come as no surprise that a CEO is not the person who actually manages the calculation of loan-loss reserves and accounting models. Depending on the structure of the company that is more likely to be done by a Chief Financial Officer, a Chief Risk Officer or by their subordinates. There was nothing precluding Duo from asking Fairstone who had the most information about a particular issue and issuing a summons for that person to appear either on a Rule 39 examination or at trial.
257In addition, the suggestion that Mr. Wood had limited knowledge of the accounting models must also be viewed in context. This is not a situation where Mr. Wood denied all knowledge. He confessed to a limitation of his knowledge when he was being examined on detailed entries in an Excel spreadsheet over a video examination where the Excel spreadsheet was being shown to him on a shared screen. That creates a significant number of limitations on a witness who was not the person who created the Excel spreadsheet or responsible for the detailed analysis.
- The Accounting Steps Fairstone Took
258In Q1 2020 Fairstone increased its LLR by $22.5 million based on Moody’s unemployment data. This does not reflect actual deterioration in the quality of the loan portfolio but is intended to reflect expected future losses. Despite the increased LLR, Fairstone’s operating income improved by 3% over the first quarter of 2019.
259In April 2020, Duo and the Office of the Superintendent of Financial Institutions asked Fairstone to issue a revised financial forecast that reflected the expected effects of the pandemic. That forecast was based on the application of the Base LLR Model and Moody’s projected unemployment numbers. In arriving at the LLR for the April forecast, Fairstone reduced the month over month change in Moody’s unemployment projections by 35%.67 It did so because Fairstone had not been experiencing the level of defaults that it would have expected based on the historic relationship of unemployment rates to default rates. This disconnect arose because of the historically unprecedented level of government financial support for consumers and businesses in response to the pandemic. This support included the Canada Emergency Response Benefit or CERB which provided significantly more support for the unemployed than traditional unemployment insurance benefits did.
260Despite the adjustments to Moody’s unemployment projections, the April forecast nevertheless projected an increase to the LLR at the end of Q2 of approximately $35 million.
261By June, 2020, Fairstone had noticed that the default rates that had been predicted by the Base Model in the April forecast, even as adjusted with a lower unemployment rate input, were still significantly higher than Fairstone had actually experienced. As a result, Fairstone prepared a new forecast in June, 2020. The June forecast essentially mirrored what Fairstone actually reported at the end of its second quarter ending June 30, 2020. It showed a significantly higher operating income to that anticipated in the April forecast, although still less than the projected operating income under its initial plan.
262When preparing the June forecast, the Base Model continued to suggest an increase to the LLR of $35 million. The losses that Fairstone had actually experienced did not reflect anything of that magnitude.
263In addition to using the Base Model, Fairstone also had a second model that it had historically used called the Roll Rate Model. The Roll Rate Model is one that tries to estimate the rate at which loans at an early stage of default will “roll” into later stages of default and ultimately into total default or bankruptcy. By June, the Roll Rate Model was able to look at default data from March and April and compare the rate at which those defaults led to further defaults in April, May and June. The Roll Rate Model showed that the subsequent defaults in April, May and June were lower than those historically experienced and would result in lower projected defaults going forward.
264The Roll Rate Model that Fairstone ran in June suggested that Fairstone reduce its LLR by $10 million. Doing that would have increased Duo’s purchase price by $10 million.
265Fairstone neither increased nor decreased its LLR in the June forecast or at the end of the second quarter. Instead, it held its LLR flat. In doing so, it applied management judgement to further dampen the effect of the effect of unemployment rates predicted by the Base Model but did not go so far as to adopt the LLR that the Roll Rate Model suggested.
266In my view, this did not amount to a change to the accounting methods, principles, classifications or practices that Fairstone used before signing the SPA.
267Fairstone had historically used both the Base LLR Model and the Roll Rate Model. As a general rule, Fairstone compared the projections that the Base LLR Model generated to those of the Roll Rate Model and, if appropriate, made adjustments to arrive at the actual reserve taken for a given period.
268Using a Roll Rate Model and making adjustments to a Base Model as a result was, on the evidence before me, an ordinary occurrence and something that Duo knew about before signing the SPA.
269Duo’s expert, Dr. Sabry, admitted that Roll Rate Models are standard tools that lenders use to project losses in their portfolios. .
270PWC noted in its due diligence report that Fairstone used Base Models to estimate loan-loss reserves but that Management adjusted the models’ results by applying their own judgment where they felt it appropriate.
271Moreover, the application of management judgement to the Fairstone Base Models was consistent with what PWC recommended should be done.
272It appears that Fairstone was doing precisely what PWC told Duo should be done.
273Actual results in Q2 appear to vindicate the steps Fairstone took. By way of example, Moody’s unemployment forecast seemed overly pessimistic. In April, Moody’s predicted rate of unemployment in Canada was 15.7%. The actual rate was 13.7%. Fairstone’s projected loan write-offs in the April forecast were 8.2%. Actual write offs were 6.5%. Net credit losses were 22% better-than-expected. Q2 operating expenses were 8% lower than expected.
274Increased reliance on the Roll Rate Model also appears to be empirically justified. Unemployment rates rose from 6% in February 2020 to 14% in May 2020 yet consumer insolvencies showed a dramatic decrease. As noted earlier, the credit quality of Fairstone’s portfolio appears to have improved during the second quarter despite the rapid rise in unemployment. Short, mid and long-term loan delinquencies were lower in May and June than they were in the 12 months before March 2020. Net charge-offs of loans were down.
275Fairstone’s approach to the calculation of Q2 LLRs was reviewed by its auditors, Ernst & Young. EY’s notes included the following observations about the rationale for Fairstone’s adjustments to the loan-loss reserves:
Moody’s unemployment rates were generally more pessimistic than those of the banks
Such pessimism may be unwarranted in light of emerging information on the economic recovery.
The standard quarterly LLR process compares the Base LLR Model to the Roll Rate Model.
The Roll Rate Model captures the actual lower than average delinquencies and bankruptcies.
276EY also highlighted the use of the Roll Rate Model and described it as follows:
Roll Rate Methodology – Given the unprecedent [sic] nature of the recession, both in terms of speed and severity of contraction, management assessed that the historical relationship between portfolio losses and the unemployment rate no longer remains reliable as a basis for forecasting future losses. However, the roll rate methodology captures the impact of continued lower than average delinquencies and bankruptcies on portfolio losses and the emergence/timing of those losses.
- Expert Evidence on Accounting Methods
277Duo filed an expert report by Richard Lee of Alix Partners. Mr. Lee is, among other things, a Certified Public Accountant and has over 23 years experience in accounting matters. He has served as an expert before American Federal and State courts, the UK High Court domestic and international arbitrations and has also served as an arbitrator in matters involving accounting, valuation and damage related disputes.
278Mr. Lee opined that:
Fairstone effectively shifted its methodology for determining its loan loss reserve (“LLR”) as of June 30, 2020 from primarily relying on Fairstone’s Base LLR Model (the “Base LLR Model”) to primarily relying on its Roll Rate Model (the “Roll Rate Model”).
279In Mr. Lee’s view, until Q2 2020, Fairstone had essentially relied upon the Base Model to calculate LLRs and applied some adjustments as suggested by the Roll Rate Model. In Q2 2020, it did the opposite and relied primarily on its Roll Rate Model.
280A change to the weight one ascribes to various discretionary factors that go into a decision is not, in my view, a change in “accounting methods, principles classifications or practices.” Fairstone applied the same accounting method and practices in the June forecast and in the second quarter as it had done historically. It simply ascribed more weight to the Roll Rate Model then to the Base Model. Even then, it did not adopt the Roll Rate Model but applied judgement to reconcile the two. Doing so was in line with what a number of independent accounting sources have recommended and as noted in paragraph 271 above, was in line with what PWC recommended.
281The thrust of Mr. Lee’s evidence was to compare the size of adjustments that Fairstone made using the Roll Rate Model in Q2 2020 to the size of adjustments it made in other periods. By way of example, Mr. Lee noted that the Q2 2020 roll rate adjustment was 30 times higher than the totality of roll rate adjustments in 2018 and 2019. While that number sounds dramatic at first, like most numerical comparisons, it calls for closer examination to determine the reason for the difference in magnitude. In the circumstances of this case, there are good reasons to explain why it was so much higher. Previous economic contractions had not been accompanied by massive increases in government support for people who had lost employment or had suffered business losses. The Base Model simply generated a number by applying the historic relationship between unemployment rates and loan defaults. In the face of massive government aid, that historical relationship was no longer accurate.
282The Roll Rate Model predicted what future losses may be based on past losses. That model demonstrated that the payment defaults in March and April did not turn into longer-term payment defaults in April May and June. As a result, projected losses were significantly lower than those historically associated with the Base Model.
283Mr. Lee also had limited qualifications for the task at hand. Fairstone’s accounting is governed by International Financial Reporting Standard 9 (IFRS 9). It came into effect in Canada on January 1, 2018. It has not been not adopted in the United States. Mr. Lee’s audit mandates were by and large American and preceded IFRS. Mr. Lee conceded that he was not a specialist in IFRS and had to consult with others if an engagement involved IFRS.
284Mr. Lee conceded in cross-examination that he was not suggesting that it was as improper for Fairstone to make the Q2 2020 adjustment nor was he providing an opinion on whether the approach Fairstone took in Q2 complied with IFRS 9 or was required by IFRS 9. He was merely identifying what Fairstone did and quantifying it.
285That is not a particularly helpful exercise for the court. All agree that IFRS requires management to apply judgement. The fact that management may have applied its judgement differently in the second quarter of 2020 than it had in other periods comes as no surprise. That is inherent in the exercise of judgement. It is not applied in an identical mechanical manner to each circumstance but it is calibrated to the facts that hand.
286Fairstone delivered an expert report from Catherine Tremblay of MNP LLP. Ms. Tremblay has, among other things, been a Chartered Professional Accountant since 1996. Like Mr. Lee, she has numerous academic credentials published articles, speaking engagements and transactional experience. She notes that Mr. Lee does not consider the requirement under IFRS to apply management judgement to develop accounting estimates and projections. In Ms. Tremblay’s view it was appropriate for management to consider the information it had which suggested that the Base LLR Model was no longer accurate.
287Duo asked me to discount Ms. Tremblay’s evidence. In support of that submission it cited three other cases in which her opinions had not been accepted, one of which criticized her report. I must base my assessment of Ms. Tremblay’s evidence on the quality of her evidence in this case, not on what other courts found in other cases. Ms. Tremblay struck me as a candid, reliable and credible witness. She readily conceded points against her. She was not defensive or argumentative even in the face of fairly aggressive cross-examination, particularly on the issue of courts that had not accepted her evidence. I found her to be consistently matter-of-fact and someone who was trying to present objective evidence rather than advance any particular position.
288As already noted, Fairstone’s Q2 LLRs were reviewed by EY. Much was made at trial about the nature of EY’s engagement. In cross-examination of Ms. Tremblay and in closing argument, Duo heavily criticized her reliance on the EY review. Duo submits that Ms. Tremblay would not concede that EY could be mistaken in its views.
289I do not think that is a fair characterization of Ms. Tremblay’s evidence. Ms. Tremblay pointed to the standards that govern an auditor’s review of interim financial statements 68and noted that it requires an auditor to perform such additional procedures as are necessary to resolve any concern of which the auditor becomes aware that may represent “a possible misstatement that is not clearly trivial.”
290In the absence of any evidence to the contrary, Ms. Tremblay assumed that EY was acting professionally and in accordance with the applicable accounting standards. That is a fair assumption to make. People are presumed to be acting appropriately unless there is evidence to the contrary. EY, like anyone else before a court, is entitled to that presumption.
291Moreover, EY did not merely accept management’s view but subjected it to assessment. In this regard, EY noted:
As part of our review procedures, we:
Held extensive discussions with management to understand the process applied inputs used to update the impact of COVID – 19 for Q2 2020. Significant judgements and assumptions were made in the evaluation of the unprecedented nature of the recession (induced by the COVID – 19 pandemic) on the LLR, which already involves extensive use of assumptions and judgement under more normal circumstances;
Reviewed the roll rate methodology and challenged management assumptions (the main assumption being the period under which delinquencies and bankruptcies will remain at a lower rate before going back to normal once the government incentive (sic) decrease during the fall); and
Reviewed Moody’s publications used by management to assess the macroeconomic environment and its impact on the LLR and compared Moody’s economic variable (sic) to other financial institutions in the Canadian market.
Based on our review procedures, we believe that management’s position at June 30, 2020, is adequate.
292The fact that IFRS required a significant degree of management judgement also comes as no surprise. All parties agreed that IFRS was less prescriptive than GAAP.
293As Ms. Tremblay’s report notes, the International Accounting Standards Board published a specific document providing guidance on the accounting applicable to credit losses in light of the Covid – 19 pandemic in which it states:
IFRS 9 sets out a framework for determining the amount of expected credit losses (ECL) that should be recognised. It requires that lifetime ECLs be recognised when there is a significant increase in credit risk (SICR) on a financial instrument. However, it does not set bright lines or a mechanistic approach to determining when lifetime losses are required to be recognised. Nor does it dictate the exact basis on which entities should determine forward looking scenarios to consider when estimating ECLs.
IFRS 9 requires the application of judgment and both requires and allows entities to adjust their approach to determining ECLs in different circumstances. A number of assumptions and linkages underlying the way ECLs have been implemented to date may no longer hold in the current environment. Entities should not continue to apply their existing ECL methodology mechanically. For example, the extension of payment holidays to all borrowers in particular classes of financial instruments should not automatically result in all those instruments being considered to have suffered an SICR.
Entities are required to develop estimates based on the best available information about past events, current conditions and forecasts of economic conditions. In assessing forecast conditions, consideration should be given both to the effects of COVID – 19 and the significant government support measures being undertaken.
294IFRS 9 requires the application of management judgement at all times. IFRS issued a particular reminder underscoring the need to apply judgement to models in order to reflect the impact of government support measures. Fairstone was legally required to publish accounting forecasts and statements that complied with IFRS 9. To have failed to apply judgement would not be operating in the ordinary course. The ordinary course was to apply judgement. That is exactly what Fairstone continued to do. To have applied the Base Model mechanically would have been to cease to operate in the ordinary course.
295Finally, any moral hazard or other risk to which Duo might be exposed as a result of Fairstone’s estimates of LLR’s in the second quarter of 2020 is addressed by article 2.7 of the SPA. Recall that the purchase price was based on Tangible Shareholders Equity plus a premium. The calculation of the LLR is a component of Tangible Shareholders Equity. Any disputes about the value of Tangible Shareholders Equity for purposes of calculating the purchase price on closing are to be determined by KPMG pursuant to a detailed, five-page procedure outlined under section 2.7 of the SPA. Duo is therefore not without protection against what it alleges are improper LLR calculations. It strikes me that KPMG is better equipped to form views about accounting judgements and LLR calculations than a court is.
E. Consent
296Section 5.1 of the SPA allows Fairstone to operate outside of the ordinary course if it has obtained the prior written consent of Duo which consent shall not be unreasonably withheld.
297Fairstone did not seek consent because it took the position it was operating within the ordinary course.69 I agree. However, if Fairstone had been operating outside of the ordinary course it would have been unreasonable for Duo to withhold consent.
298Duo submits that Fairstone’s failure to seek consent is the end of the analysis and that I am not entitled to go any further to inquire whether Duo should have consented to those steps or whether it would have been unreasonable for Duo to refuse consent. In support of this proposition, Duo relies on the unreported oral reasons of the English Court of Appeal in Hendry v. Chartsearch.70 That case is distinguishable. In Hendry, the plaintiff was the assignee of a contract. The original contract was between Interface and the defendant Chartsearch pursuant to which Interface agreed to provide data processing services to Chartsearch. After the relationship between Chartsearch and Interface had broken down and after Chartsearch had taken the position that the contract was at an end, Interface purported to assign the contract to the plaintiff so that the plaintiff could sue Chartsearch. Interface assigned the contract without seeking Chartsearch’s consent.
299Consent to the assignment of a longer-term relational contract for the provision of services raises far different questions than does the consent to the steps Fairstone took in response to the pandemic. The steps Fairstone took are specific and discrete. The reasonability of withholding consent on the facts before me is far easier to determine than in a long-term or relational contract. Moreover, the consent issue I am being asked to decide on did not arise after the contract was terminated but while it was alive. Finally, in the case at hand consent is not being sought for the purpose of suing a party but for the purpose of maintaining the SPA and the underlying business.
300By incorporating a reasonableness standard into the SPA, the parties clearly agreed that someone would evaluate the issue in the event of disagreement between the parties.
301Duo submits that the reasonable consent required Fairstone to give Duo a list of the alternatives Fairstone considered before taking the steps it did. According to Duo, if Fairstone were implementing only the most costly solutions, Duo should know about alternatives that were considered but rejected. While that proposition might be theoretically correct in an appropriate case, it is not pertinent here.
302The steps at issue were not particularly complex, were often mandated by government regulation, were common to those that businesses around the world had taken, were of short to midterm duration and were likely similar to the steps Duo had taken. Although Duo was in a similar business to that of Fairstone, Duo never suggested any alternatives that were preferable to the ones Fairstone implemented, either before or during trial. Finally, at trial Duo did not introduce any evidence of what it would have done differently or what it did differently in its own business.
303In the foregoing circumstances I am satisfied that, to the extent that steps Fairstone took were outside of the ordinary course of business, Duo would have been unreasonable to withhold consent.
IV. Amortization Event
304Section 6.2 (a) of the SPA makes it a condition of closing that:
… no condition, development, fact, effect, circumstance, change or event shall exist that results or would reasonably be expected to result in an [ ] Amortization Event under any Financing Document.
305The Financing Documents referred to in s 6.2 (a) are those that relate to the notes Fairstone issued in connection with its loan portfolios. Fairstone does not fund its consumer lending with its own capital. Instead, it borrows from institutional investors by securitizing its loan portfolios and issuing notes to investors in respect of each individual portfolio.
306Fairstone has five securitized portfolios. Two are at issue here. They have been referred to as the 2019 – 1 and the 2019 – 2 portfolios. The former has approximately $322.4 million in notes outstanding. The latter has approximately $1.377 billion in notes outstanding.
307The financing documents associated with each of the portfolios contain what are called Amortization Events. Amortization Events are designed to protect the noteholders. If an Amortization Event occurs, it results in a redirection of cash flow from the portfolio. In the absence of an Amortization Event, Fairstone’s only obligation to the noteholders is to pay interest. Capital does not have to be repaid until the maturity of the notes. As a result, Fairstone enjoys the benefit of all cash in excess of the interest rate payable to noteholders and certain expenses.
308If an Amortization Event occurs, cashflow from the payments on the underlying loans is directed exclusively to noteholders. In addition, after an Amortization Event the noteholders can no longer be called on to provide further funding for the facility. Fairstone would therefore be unable to underwrite any new loans without alternative sources of capital.
309This creates the potential for a longer-term effect which could work to the detriment of a purchaser.
310Duo alleges that two Amortization Events are reasonably expected to occur: an average loss ratio trigger and a note balance event. An average loss ratio trigger occurs if losses within a portfolio exceed a certain percentage. A note balance event will occur on March 22, 2021 if Fairstone is not able to pay off the $322.4 million owing on the series 2019 – 1 notes by then.
311Duo’s letter of May 27, 2020 also alleged an Amortization Event because of a pool balance deficiency. This ground was not pursued at trial.
312All parties agree that an Amortization Event had not occurred by the time of trial. As a result, the Amortization Event issue turns on whether an Amortization Event “would reasonably be expected to result.” Duo submits that since the Amortization Event condition was inserted for its benefit, if an Amortization Event would reasonably be expected during the life of either of the facilities, the closing condition has not been satisfied and Duo is not required to close. Fairstone submits that, if an Amortization Event has not occurred by the time of closing, that is the end of the analysis and Duo must close.
313I have concluded that, on the record before me, and Amortization Event would not be reasonably expected to result. I will examine the issue by considering each of the alleged Amortization Events and the burden of proof related to those alleged events.
A. The Average Loss Ratio Trigger
314The resolution of both alleged Amortization Events turns heavily on an assessment of the evidence of Duo’s expert Dr. Faten Sabry and Fairstone’s expert Linda Roy.
i. Evidence of Dr. Faten Sabry
315Dr. Sabry is highly credentialed. She is an economist, a statistician, has been a post-doctoral fellow, an assistant professor and is currently the head of the securities and finance practice of NERA Economic Consulting. She has experience in forecasting loan losses and was responsible for allocating an $8.5 billion settlement between Bank of America and investors in approximately 530 securitized loan portfolios based on the expected lifetime losses in each portfolio.
316Unlike Fairstone’s expert, Dr. Sabry did not provide an opinion about whether an Amortization Event was reasonably expected to occur. The questions she answered were substantially more nuanced. As she set out in her affidavit, she was asked:
(i) Whether Fairstone’s contemporaneous analyses and their estimates of the future losses of the loan portfolio are consistent with the conclusions of Linda Roy, in her report dated July 15, 2020; and
(ii) Whether contemporaneous expectations about unemployment given the Covid – 19 pandemic and its impact on the forecast of the future losses of the loan portfolio would have likely resulted in an Amortization Event in respect of either FFIT I 2019-1 or FFIT I 2019-2.
317More plainly put, Dr. Sabry’s question (i) is whether Ms. Roy’s report is consistent with the results generated when using Fairstone’s Base LLR Model and Moody’s unemployment rates. Her question (ii) simply plugs different unemployment rates into the Fairstone Base LLR Model to show at what points Amortization Events are projected to occur when using these different unemployment rates.
318The first question may have been motivated by a misunderstanding of what Ms. Roy was doing. Dr. Sabry was apparently not aware that Ms. Roy and Fairstone had not been using Moody’s unemployment projections but had used an unemployment projection which discounted Moody’s number and further dampen the effect of unemployment. Even after Dr. Sabry received a reply report from Ms. Roy which made the difference clear, she did not revise her question but provided a critique of Ms. Roy’s reply report.
319Dr Sabry reruns Fairstone’s Base LLR Model and concludes that when it is run without adjustments it would have predicted an Amortization Event in the 2019-1 Series in August and September 2020 and between December 2020 and March 2021. The same approach to the 2019 – 2 Series would have resulted in Amortization Event between July 2020 and January 2022.
320In addition, Dr. Sabry ran the Base LLR Model using various inputs to demonstrate its sensitivity to only minor changes. By way of example, she notes that if unemployment estimates in the March 2020 model are increased by only 0.08 percentage points, the model projects an Amortization Event in January 2021.
321While these analyses are perhaps interesting, they are divorced from reality. As of May 27, 2020, the loss trigger ratio in the 2019 – 1 facility was 13.85%. Actual losses in the portfolio stood at 8.2%. Between May 2019 and January 2020 losses had ranged between 3.8% and 9.8%.
322The loss trigger ratio in the 2019 – 2 facility as of May 27, 2020 was 12.09%. Actual losses stood at 7.2%. That facility only began operating October 2019. Its loss ratios had ranged from a low of .8% November to a high of 8% in April 2020. As noted earlier, however, the lower losses in the early months were attributable to the manner in which Fairstone records losses during the first six months of a portfolio and do not reflect actual historical results.
323Both sides agree that no Amortization Event occurred to the end of trial even though Dr. Sabry reported that the application of unadjusted numbers would have projected Amortization Events beginning in August for the 2019-1 portfolio and in July for the 2019-2 portfolio.
324Dr. Sabry’s models in effect continue the use of the Base LLR Model to project an Amortization Event. Her projections of what would happen compared with what actually did happen for the two securitized facilities are set out in the two tables below. The upper, solid horizontal line represents the trigger ratio. The lower, solid inclined line shows what actually occurred. The lower dotted line shows Fairstone’s Q1 forecast. The upper dotted line shows Dr. Sabry’s forecast. It is readily evident from these figures that Fairstone’s Q1 forecast was not realistic. Actual results were far more positive than the Q1 forecast suggested. This demonstrates the need to revise the historic models on which the Q1 forecast was based and supports Fairstone’s decision to do so. A forecast that departs entirely from the reality of the situation is not much of a forecast and certainly not useful to project what is likely to happen.
325The two figures below show Fairstone’s Q2 forecast. Again, the upper, solid horizontal line reflects the trigger ratio. The solid lower line reflects actual results. The dotted middle line reflects Fairstone’s projections and the dotted upper line reflects Dr. Sabry’s projections.
326Once again, Dr. Sabry’s projections did not materialize. In both cases, Dr. Sabry’s models predicted an Amortization Event before the hearing of this matter. All parties agree that never happened.
327Dr. Sabry’s modelling continued to predict Amortization Events in July and August 2020 even though those did not occur. Dr. Sabry’s modelling consistently showed loss ratios higher than those actually recorded. She explained that the purpose of her report was to show what would happen if Fairstone’s Base Model were used without management adjustments. If anything, that underscores the importance and relevance of the management adjustments. Even with Fairstone management’s adjustments, the projections were still more conservative than actual results as the tables above demonstrate.
328Apart from what I view as the diminished utility of the exercise in which Dr. Sabry engaged, the degree of her personal experience with issues like the ones at hand was significantly less relevant than the experience of Ms. Roy. As I will explore in more detail below, Ms. Roy assessed the likelihood of an Amortization Event by evaluating the quality of the portfolios themselves. This is a skill she developed over 30 years during her tenure at National Bank. Dr. Sabry fairly conceded that she had no expertise in estimating losses, developing credit scores, determining when consumer loans should be refinanced or otherwise managing Canadian consumer loan portfolios.
329The lack of Canadian experience has some relevance here because, as Duo’s own investment thesis for the Fairstone purchase demonstrated, a Canadian low wage earner has approximately $9,000 more disposable income than an American low wage earner because of Canada’s superior social safety net. This was one factor that made Fairstone an attractive target for Duo. It would also reasonably lead to a different default profile in the portfolios. Ms. Roy actually assessed the likelihood of default in the portfolios. Dr. Sabry merely took different unemployment numbers and plugged them into Fairstone’s Base Model to determine whether and when particular unemployment rates would trigger Amortization Events.
ii. Evidence of Linda Roy
330Fairstone’s expert on the Amortization Event issue was Linda Roy. She tracked the language of the SPA and provided an opinion on whether there has been an Amortization Event or whether there was any condition, development, fact, effect, circumstance, change or event that results or would reasonably be expected to result in an Amortization Event. She concluded there was not.
331Ms. Roy was employed by National Bank for 35 years. She was its Managing Director of Fixed Income and Securitization between 2006 and 2019 when she retired. In that capacity she maintained an overview of the market, was responsible for any exposure the bank would take, and tracked performance and amortization triggers of a wide variety of securitized portfolios. She received monthly performance reports for all debtors which she analysed by way of her own personal financial dashboard to help follow trends. She was instrumental in leading the bank through the 2007– 2009 financial crisis.
332Ms. Roy criticizes Dr. Sabry’s approach because it ignores the effect of government stimulus on the portfolios. A number of independent sources confirm that old forecast models no longer work given the level of government stimulus in response to the pandemic and that adjustments to those forecasts need to be made. These include:
KPMG stating:
Existing ECL71 models will use historical experience to derive links between changes in economic conditions and customer behavior, and ECL parameters such as loss rates, probabilities of default and loss given default. However, these historical relationships are unlikely to read across to the COVID – 19 pandemic. Therefore, adjustments to model results, based on expert credit judgement, could be necessary to reflect the information available at the reporting date appropriately.
IFRS stating:
A number of assumptions and linkages underlying the way ECLs have been implemented to date may no longer hold in the current environment. Entities should not continue to apply their existing ECL methodology mechanically.
GoEasy Q1 2020 financial statements saying:
Management also applies expert credit judgment in the determination of ECL which is informed through the analysis of relevant historical experience, multiple stress-weighted scenario analysis and the consideration of the significant recent government stimulus measures, changes in the Company’s policies and procedures and the level of creditor insurance within the portfolio.
333In response, Duo notes that the CERB was scheduled to end in early October 2020. While true, Fairstone replies that three new government programs were announced to compensate for the wind down of the CERB.
334The fundamental difference between the evidence of Ms. Roy and that of Dr. Sabry is that Dr. Sabry relied entirely on unadjusted Base Model forecasts. While forecasts were of interest to Ms. Roy, she placed far greater weight on the underlying credit quality of the loan portfolios. Ms. Roy takes this approach because, in her view, just because a forecast predicts an Amortization Event, does not mean that it will occur. Indeed, as Ms. Roy explained, the whole purpose of obtaining projections is to give management a tool that it can use to avoid Amortization Events, not to run headlong into them. If the projection is accurate, it predicts what will happen if management makes no refinements to its portfolios.
335Ms. Roy, likened this to driving a car slowly, knowing that there is a cliff several hundred kilometres down the road. The purpose of knowing that there is a cliff down the road is not to drive off the cliff but to avoid it.
336That is precisely how the models are intended to be used by management: to help them change course to permit them to avoid the projected Amortization Event. This can be done in many ways including changing the mix of loans in a portfolio and refinancing loan balances through customer engagement.
337By way of example, losses on secured loans are approximately 2% a much lower rate than losses on unsecured loans. If an Amortization Event is projected, management can lower its interest rate on secured loans to make them more attractive and increase the percentage of secured loans within a particular portfolio thereby avoiding potential Amortization Events. Similarly, increased customer engagement involves speaking with customers who may be in a more precarious situation to help them manage the situation before a default occurs. This can be done through payment deferrals or refinancing that extends the term of the loan but results in lower monthly payments which the customer can maintain more easily.
338In determining whether an Amortization Event existed or was reasonably expected to exist, Ms. Roy analysed a wide variety of parameters for the Fairstone portfolios including historical data, loan balance data, the weighted average credit score of the individual debtors within each portfolio and the average loss ratio broken down between loan types and maturity dates of individual loans. By way of example, Ms. Roy noted that the weighted average life of the loans in the Fairstone portfolios is between 1.5 and 2 years. This means that when Fairstone changes its underwriting criteria, it influences the performance of the portfolio more quickly
339I found Ms. Roy’s explanations to be compelling and persuasive. Those explanations were unshaken during cross-examination.
340Fairstone’s actual results during the pandemic demonstrate that the combined effect of government support and effective management have actually improved credit quality within the portfolios.
341By way of example, in May, .75% of its loans were in an early stage of delinquency (30 – 59 days). By June the figure was .66%. This compares to a figure of .92% in the 12 months preceding March 2020. Early-stage delinquency is an important factor to consider because it reflects the number of write offs one is likely to experience in the future. Fairstone’s stage II delinquencies (60 to 89 days) experienced a similar decrease. In March they stood at .6%. In June they stood at .51%. This compares to a figure of .63% in the 12 months preceding March 2020.
342Bankruptcies also decreased in portfolios. Bankruptcies decreased by 40.8% between March and April 2020 and by a further 20.1% between April and May 2020. Although bankruptcies increased by 10.7% in June, 2020, the year-over-year bankruptcy rate was still 30.8% less in June 2020 than June 2019.
343In response to the pandemic, Fairstone has been accumulating cash, tightening lending requirements and has engaged in a high level of customer interaction to avoid loan delinquencies. By way of example, Fairstone increased its cash from $150 million in February to $213 million as of June 30, 2020. Fairstone has used its cash to reduce its debt by $188 million in Q2 2020.
344As noted earlier, Fairstone’s particular skill in managing through economic downturns was one of the reasons Duo found it so attractive.
B. Note Balance Event
345As stated above, a note balance event will occur on March 22, 2021 if Fairstone is not able to pay off the $322.4 million owing on the series 2019 – 1 notes by then. Duo alleges that Fairstone will be unable to access funds to pay off the 2019 – 1 facility and that an Amortization Event is therefore reasonably expected for that portfolio.
346Duo bases this submission on the fact that Fairstone received a B1 rating from Moody’s and that B ratings are reserved for investments subject to high credit risk.
347This caused me further concern about Dr. Sabry’s evidence. Dr. Sabry based her concern about Fairstone’s ability to refinance on Fairstone’s corporate credit rating. The notes themselves, however, are rated separately from Fairstone. The senior notes have a AAA rating, the highest rating possible. When confronted with this in cross-examination, Dr. Sabry explained that the credit rating of Fairstone was nevertheless relevant because some of the Amortization Events related to Fairstone’s ability to continue in business such as, for example the ability to continue to collect on outstanding loans. Dr. Sabry had to concede, however, that those factors were also taken into account in the AAA rating that notes received. As a result, the reference to a B1 rating leaves a misleading impression. This was somewhat of a consistent theme with Duo’s experts. I had the uncomfortable impression that their goal was to create arguments to advance their client’s case rather than to provide objective opinions about the matters at issue.
348Next Duo argues that there is no evidence that Fairstone will be able to access the funds needed to pay off the notes. To support this view, Duo introduces evidence from Dr. Sabry pointing to a Bank of Canada survey of approximately 100 businesses, a majority of which indicate that they believe financing has become more difficult. Dr. Sabry also points out that credit spreads, the premiums investors require to invest, have increased. That is insufficient evidence of an inability to refinance approximately 10 months from May 27, 2020. Credit spreads fluctuate regularly. The fact that credit spreads have increased does not mean Fairstone cannot obtain financing. Recall that Fairstone managed its way through the 2008 financial crisis when credit had completely locked up and was simply not available. Fairstone has already begun preparing for that eventuality by accumulating cash and paying down debt.
349Ms. Roy notes that a number of securitization financing transactions have closed during the pandemic which demonstrates that finance markets remain open. She also notes that Fairstone’s recent transactions were over subscribed by sevenfold. Finally, the financiers under the 2019 – 2 portfolio have committed to provide approximately $500 million more in financing. If Fairstone needed to, it could use those funds to retire the 2019 – 1 facility.
C. Burden of Proof
350As noted, no Amortization Event had occurred by May 27, 2020, August 14, 2020, or indeed by the time of trial. The issue is whether a condition existed as of May 27, 2020 or August 14, 2020 that “would reasonably be expected to result” in an Amortization Event.
351Fairstone submits that the Amortization Event must exist at the time of closing (or August 14, 2020) on a balance of probabilities for Duo to avoid closing.
352Duo submits that since the Amortization Event condition existed solely for its benefit, if, at closing, it reasonably expected an Amortization Event at any point during the lifetime of the facilities, it was entitled not to close. According to Duo, this is so even if it may also have been equally reasonable not to expect an Amortization Event.
353Neither submission strikes me as the appropriate answer.
354Fairstone’s submission does not give sufficient weight to the wording of the agreement. The SPA does not require that no Amortization Event exist at the time of closing. Rather, it says that no condition shall exist at the time of closing which condition “results or would reasonably be expected to result in an Amortization Event.” What must exist at closing is not the Amortization Event but a condition. The plain wording of the agreement makes clear that the actual Amortization Event can arise after closing.
355At the same time, Duo’s submission runs afoul of two fundamental contractual principles. First Duo’s emphasis on its own subjective belief runs afoul of the principle that contracts are to be interpreted objectively. Duo’s submission that it can avoid the contract if it reasonably expected an Amortization Event to occur loses all sense of objectivity and turns the issue into a purely subjective inquiry into Duo’s expectations.
356Second, extending the horizon of the Amortization Event over the life of the facility gives insufficient weight to the general principle that systemic risks arising after closing belong to the purchaser. The Amortization Events discussed at trial all arose out of systemic risks such as the pandemic and its effect on borrowers’ ability to repay debt.
357The 2019 – 2 facility expires on October 31, 2022, more than two years after the trial occurred. This makes Duo’s argument even more subjective. It would allow Duo to avoid the contract if it could develop a reasoned argument to the effect that an economic downturn within two years of closing would lead to an Amortization Event. This gives far too little weight to the concept that the SPA was a contract that was meant to be enforceable, not an option that Duo was free to exercise or not as it pleased on the closing date.
358While I do not think it is wise to try to define the precise timeframe in which an Amortization Event is to be expected, I return to the Supreme Court of Canada’s language that I cited earlier when addressing similar language in connection with MAEs. The Supreme Court required the expectation to be likely not merely possible. The expectation must be “tethered to the realities, not the possibilities.” When I apply those concepts to Duo’s expectation of an Amortization Event, I conclude that Duo has not met its burden of proof.
359Duo’s argument that an Amortization Event is reasonably expected when using the Base LLR Model is not tethered to realities because it does not reflect the need to apply management judgement to the base models and gives no weight to the concept that such models are designed to help management avoid Amortization Events, not to drive headlong into them. If projections predict an actual or proximate Amortization Events, what would more reasonably be expected than and Amortization Event is that Fairstone’s management would employ the panoply of tools it has successfully employed over the past 100 years to navigate the business away from the Amortization Event. In the circumstances of this case, this expectation is not naïve, blind optimism in management’s self-aggrandizing statements, for which a court would have little time in any event. I am relying on Duo’s views in this regard. It was Duo that recognized the strength of Fairstone’s management and viewed its ability to navigate economic difficulties successfully as one of Fairstone’s prime assets.
360To date, none of Fairstone’s lenders have expressed any concern about Amortization Events being expected. The securitizations are rated by two external rating agencies: Moody’s and DBRS. Neither had changed the rating or issued any sort of a ratings watch or warning.
361On the evidence before me, since the SPA was signed, Fairstone management has demonstrated a fair-minded, responsible attitude. This is not a case of a self-interested, opportunistic management group seeking to extract the maximum purchase price from an unwitting purchaser.
V. Access Covenant
362Section 5.3 of the SPA provides that Fairstone shall:
…furnish to Purchaser… such financial and operating data… within a reasonable time following receipt of a request [as] is reasonably necessary for the consummation of the transactions…
363Duo alleges that Fairstone breached its covenant to provide Duo with all information reasonably necessary for closing. Fairstone denies any breach.
364As a starting point it is worth noting that in its letter of May 27, 2020, Duo did not raise the denial of information as a grounds for refusing to close. It appears that, up to that point, Duo was content with the information it received.
365Many of the requests arose after April 1, that is to say after Duo had already taken the position that Fairstone was in material breach of the agreement. Fairstone nevertheless continued to respond. Fairstone became concerned that the information requests were designed to overwhelm Fairstone employees and amounted to a fishing expedition aimed at providing Duo with arguments about why it need not close.
366Duo made thousands of information requests. Approximately 36 Fairstone employees were involved in providing answers. At the end of all of that, Duo alleges that nine requests were unanswered and seven were insufficiently answered.
367I can readily understand frustration on both sides. On the one hand, Duo had a right to information and wanted to get it quickly. On the other, Fairstone while obliged to provide information, also had a business to run in the particularly trying circumstances of the pandemic. No doubt working remotely made information more cumbersome to track down.
368I tend to the view, however, that many of the requests amounted to more of a fishing expedition to support Duo’s position that it need not close than legitimate requests for information that Duo required to close the transaction. By way of example, many of the information requests about which Duo complained at trial related to Fairstone’s failure to provide information about alternatives it had considered to the steps it took in response to the pandemic. Given that Duo was in a business very similar to that of Fairstone’s, Duo did not really require a list of every alternative Fairstone had considered. If Duo was of the view that Fairstone was doing something inadvisable, Duo had enough information from its own business operations to engage in a discussion to arrive at a more acceptable alternative. There was no evidence that Duo engaged in such discussions.
369At the end of its written argument, Duo attached as Annex 1, an additional 17 page, single spaced chart setting out and explaining the need for the information requests. No witness went to any of the specific requests to demonstrate why Duo legitimately required any of them. Although Duo’s affiant Ms. Fahie addressed the information concern generally in her affidavits, she does not address the specific requests that Duo says remained unanswered at trial and does not specifically say why Duo required that to close the transaction.
370By way of example, one of the unanswered requests about which Duo complains is set out as item 2 (a) in its written argument and asks for:
Projections as to expected deferment activity, AOTs, and RBOs, and any other loss mitigation or modification activities by product within the next 12 months on an LTM basis, in each case with projections broken down by COVID – 19 Program Deferment, non-COVID – 19 deferment or other mitigation/modification.72
371Item 5 (a) reads:
Production by product, presented in terms of production in-branch in-person, production in-branch by phone and online loan fulfillment, and identifying for each category both loan renewal and new loan origination.
372It is difficult to understand why such microscopic detail was required to close. Duo says the requests would provide information about potential breaches of the MAE, Ordinary Course and Amortization Event covenants. Given that I have already found that these covenants were not breached, the requests are a bit of a red herring.73
373A further red herring that Duo pursued at trial was information about an amendment to the Note Purchase Agreement. It too is somewhat beside the point because the proposed agreement never proceeded.
374Finally, Duo says it has been treated unfairly because it did not receive information about management adjustments to Moody’s unemployment rate in a timely manner. This was a real-time transaction and real-time litigation. As a result, all parties were obliged to act on a real-time schedule. Fairstone disclosed that it had discounted the Moody’s unemployment rate by 35% in paragraph 197 of Mr. Woods’ affidavit of July 15, 2020. Dr. Sabry’s reports were dated July 31 and August 21, 2020. The disclosure in Mr. Woods’ affidavit should have provided adequate time to absorb the information if not by July 31, then by August 21, 2020.
VI. Conclusion and Disposition
375For the reasons set out above, I have concluded that none of the MAE, ordinary course, amortization event or access to information covenants on which Duo relies, have been breached and none provide Duo with any basis for refusing to close the transaction.
376Both parties agree that if I find in Fairstone’s favour, the appropriate award is specific performance rather than damages. I therefore order Duo to perform the SPA and close the transaction.
377There was some suggestion during closing argument that there may be some logistical issues surrounding the closing on which the parties require my direction depending on the time at which my reasons are released. I will remain seized of this matter to provide whatever further assistance the parties require to close the transaction.
378I would also ask the parties to come up with a proposed timetable to deal with the issue of costs. My preference is that the timetable be either agreed to or that I create a timetable within two weeks of release of these reasons.
379As a final closing note, I would like to extend my sincere thanks to all of the lawyers and witnesses involved in this proceeding. Although the proceeding was hard-fought under challenging circumstances, the litigation was, at least from my perspective, a model of civility. Counsel cooperated on a broad range of logistical issues. They made efficient and judicious use of my time before and during trial. They were able to get a complex piece of litigation to trial within a matter of months. I fully appreciate how difficult that is and am indebted to all who made that possible. All counsel and their support staff should be congratulated for an extraordinarily well done job.
Koehnen, J.
Released: December 2, 2020
COURT FILE NO.: CV-20-00641857-00CL
CV-20-00643629-00CL
DATE: 20201202
ONTARIO
SUPERIOR COURT OF JUSTICE
BETWEEN:
FAIRSTONE FINANCIAL HOLDINGS INC., J.C. FLOWERS IV L.P. and VP CANADA ACQUISITION, LP
Applicants
- and –
DUO BANK OF CANADA
Respondent
AND BETWEEN:
DUO BANK OF CANADA
Plaintiff
- and –
FAIRSTONE FINANCIAL HOLDINGS INC., J.C. FLOWERS IV COINVEST CANADA L.P., J.C. FLOWERS IV L.P., and VP CANADA ACQUISITION, LP
Defendants
REASONS FOR JUDGMENT
Koehnen, J.
Released: December 2, 2020
Footnotes
- At the request of the parties I have referred to certain financial parameters in these reasons by using adjectives instead of specific numbers in order to protect confidential business information.
- Defined as 11:59 PM on the last day of the month immediately preceding the month of the closing date.
- In using the acronym MAE I will assume the reader will either read the acronym as a word (pronounced “may”) or will read it as “material adverse effect”. I will therefore use the indefinite article “a” instead of “an” unless I am quoting from another source, in which case I will use whatever indefinite article the original source used.
- Cariboo Redi-Mix & Contracting Ltd. v. Barcelo, 1991 CarswellBC 2263 (BC SC) at para. 57 [Cariboo]; Frontier Oil v. Holly Corp., 2005 Del. Ch. LEXIS 57 (Del. Ct. Ch.), p. 34 [Frontier Oil]; Akorn, Inc. v Fresenius Kabi AG, 2018 Del. Ch. LEXIS 325 (Del. Ct. Ch.), p. 138 [Akorn].
- More particularly the "Effective Time" which, for ease of reference, will be referred to as closing.
- In re IBP S'Holders Litig. v. Tyson Foods, (2001) 789 A.2d 14 (Del. Ct. Ch.), p. 68 [IBP]; Frontier Oil, supra note 2 at p. 35; Hexion Specialty Chemicals, Inc. v. Huntsman Corp., (2008) 965 A.2d 715 (Del. Ct. Ch.), p. 739 [Hexion]; Grupo Hotelero Urvasco SA v. Carey Value Added SL, [2013] EWHC 1039 (Comm) at para. 364; and 9007-7876 Québec inc. c. Provigo Inc., 2004 CarswellQue 10072 (QCCA) at paras. 54 and 109 [Provigo].
- Peart v Peel (Regional Police Municipality) Police Services Board, 2006 CanLII 37566 (ON CA), 43 C.R. (6th) 175 (Ont. C.A.) para. 139; Ronald Joseph Delisle et al, Evidence: Principles and Problems, 12th ed (Toronto: Thomson Reuters, 2018) at 44.
- Akorn, supra note 2 at fn. 619. See also Provigo, supra note 5 at para. 53, applying the same principles by invoking article 2803 of the Québec Civil Code (QCCA).
- C. (R.) v. McDougall, [2008] SCC 53, 2 S.C.R. 41 at para. 40.
- 2012 SCC 3, 1 S.C.R. 23 [Merck Frosst].
- Ibid at para. 196.
- 2005 SCC 57, 2 S.C.R. 706 [Hilewitz].
- Ibid at paras. 53 and 58
- Ibid at para. 60
- Akorn, supra note 2 at p. 162.
- Ibid at p. 162; Channel Medsystems, Inc. v Bos. Sci. Corp., 2019 Del. Ch. LEXIS 1394 (Del. Ct. Ch.), p. 69 [Channel Medsystems].
- Frontier Oil, supra note 2 at fn. 209
- Akorn, supra note 2 at fn. 646.
- Ibid at p. 151.
- Channel Medsystems, supra note 15 at p. 65; Inmet Mining Corp. v. Homestake Canada Inc., 2002 BCSC 61, 99 B.C.L.R. (3d) 93 at paras. 38-43, aff’d 2003 BCCA 610, 38 B.L.R. (3d) 248 [Inmet].
- It is worth noting that, in Channel Medsystems, the court refused to use a date earlier than that upon which the the agreement actually terminated, finding the purchaser’s effort to use an earlier date was “a pretext to elide evidence unhelpful to its case…even though [it] did not have the strength of its convictions to terminate the agreement.” See Channel Medsystems, supra note 15 at p. 66. That same reasoning applies with equal force to the case at hand.
- Akorn, supra note 2 at p. 122
- IBP, supra note 5 at p. 68.
- Cariboo, supra note 2 at paras. 54, 57; Inmet (trial decision), supra note 19 at para. 128.
- Supra note 19 at para. 128.
- Extreme Venture Partners Fund I LP v. Varma, 2019 ONSC 2907, 94 B.L.R. (5th) 38 at para. 222 [Extreme Venture]; Consumers Glass Co. v. D’Aragon, 6 B.L.R. 114 (Ont HCJ) at para. 56 [Consumers]
- I say close to zero because Fairstone does write off at bankruptcies within the first six months.
- 1998 CanLII 14814 (ON SC), 44 B.L.R. (2d) 211 (Ont. Gen. Div.) [Dynacare].
- IBP, supra note 5 at p. 88; Hexion, supra note 5 at p. 738; Frontier Oil, supra note 2 at p. 34.
- Akorn, supra note 2 at p. 144; IBP, supra note 5 at p. 68.
- IBP, supra note 5 at p. 67.
- Hexion, supra note 5 at p. 738.
- IBP, supra note 5 at p. 23 and 67-68.
- While I do not refer to specific financial parameters here, the parties have asked that the specific synergies not be disclosed.
- Extreme Venture, supra note 23 at para. 222; Consumers Glass, supra note 23 at para. 60.
- Inmet BCCA., supra note 19 at para. 102.
- See, for example, Inmet (trial decision), supra note 19 at para. 29.
- Supra note 5.
- Ibid at p. 741.
- Dynacare, supra note 25 at para. 132.
- Akorn, supra note 2 at p. 112.
- Government of Canada, “Frequently asked questions - Canada emergency wage subsidy (CEWS)” (URL: https://www.canada.ca/en/revenue-agency/services/subsidy/emergency-wage-subsidy/cews-frequently-asked-questions.html); Government of Canada, “Questions and Answers on the Canada Emergency Response Benefit” (URL: https://www.canada.ca/en/services/benefits/ei/cerb-application/questions.html); Canada Mortgage and Housing Corporation, “COVID-19: CECRA for small businesses” (URL: https://www.cmhc-schl.gc.ca/en/finance-and-investing/covid19-cecra-small-business); Government of Canada, “Canada Emergency Business Account (CEBA)” (URL: https://ceba-cuec.ca/).
- Emergency Management and Civil Protection Act, R.S.O. 1990, c E.9, s. 1; O. Reg. 50/20: Declaration of Emergency; and O. Reg. 82/20: RULES FOR AREAS IN STAGE 1.
- Emergency Order No. 1 – To impose regulations requiring physical distancing within Parks and Public Squares (April 2, 2020) (URL: http://app.toronto.ca/nm/api/individual/notice/1625.do), accessed August 30, 2020.
- Akorn, supra note 2 at p. 199-202
- See paragraph 295 below.
- Stelco Inc. 2007 ONCA 483, 2007 ON A 483, 32 B.L.R. (4th) 77 at para. 89 [Stelco]
- Pacific Mobile Corp., Re, 1985 CanLII 67 (SCC), [1985] 1 S.C.R. 290 [Pacific Mobile].
- Agricultural Commodity Corp. v. Schaus Feedlots Inc., [2001] O.J. No. 2908 (Ont. S.C.J.) at para. 16.
- Lou R. Kling & Eileen T. Nugent, Negotiated Acquisitions of Companies, Subsidiaries and Subdivisions (New York: ALM Properties, Inc, 2008), at 13-19.
- Akorn, supra note 2 at p. 196 citing Lou R. Kling and Eileen T. Nugent, Negotiated Acquisitions of Companies, Subsidiaries and Subdivisions, (New York: ALM Properties, Inc, 2008), at 13-19.
- Akorn, supra note 2 at p. 196 and fn. 775
- Stelco, supra note 48 at paras. 95-100.
- Ibid at para. 99.
- The Supreme Court of Canada also noted industry practice as one measure of determining ordinary course conduct in Pacific Mobile, supra note 45 at para. 5.
- Houlden & Morawetz, Bankruptcy and Insolvency Law of Canada, 4th ed., (Toronto: Thomson Reuters Canada Limited), at F§210(2).
- 369413 Alberta Ltd. v. Pocklington, 2000 ABCA 307, 194 D.L.R. (4th) 109 [Pocklington].
- Ibid at paras. 21-22.
- Canadian Commercial Bank v. Prudential Steel Ltd., 1986 CanLII 1793 (AB KB), [1986] 75 A.R. 121 [Prudential Steel].
- Ibid at paras. 16-17.
- Ibid at paras. 5, 17, 23.
- Pocklington, supra note 55 at paras. 3, 28.
- Subject to the nuance described earlier to the effect that it is quite possible to be outside of the ordinary course even though the nature of the business is the same at closing that signing.
- Dynacare, supra note 27 at paras. 132, 165-169.
- Creston Moly Corp. v. Sattva Capital Corp., 2014 SCC 53, 2 S.C.R. 633 at para. 47.
- I say apart from the alleged accounting changes because they could potentially require Duo to pay more for the business than it might otherwise have to.
- In addition to the fact of the reduction in Moody's employment projections, Duo submits that the information was unfairly hidden from it. I do not share that view. Mr. Wood disclosed in paragraph 197 of his affidavit of July 15, 2020 that Fairstone had adjusted Moody’s unemployment rate projections by 35% when preparing the April forecast.
- Section 7060 of the Chartered Professional Accountants of Canada (CPA) Standards and Guidance Collection: Auditor Review of Interim Financial Statements.
- Subject to the letter of March 27, 2020 in which, at the request of Steven Smith, Fairstone sought consent to the employment measures referred to earlier.
- Henry v Chartsearch Ltd. (July 23, 1998), Doc. 95/0598/1 (Eng. C.A.).
- Expected credit loss
- With the acronyms having the following meanings: RBO: Refinance Balance Only; AOT: adjustment of terms; and LTM: last-twelve-months basis.
- I underscore here that answers to those questions would not have affected my findings on the MAE, ordinary course and amortization event covenants. The requests went to the size of refinancings, adjustments or online loan origination. As noted earlier in these reasons, all of those elements had been part of Fairstone's business plan before signing. The fact that they might have played a larger role for a short period of time during the pandemic does not transform them into MAEs, out of ordinary course conduct or amortization events.