Supreme Court Judgments

Transport North American Express Inc. v. New Solutions Financial Corp., [2004] 1 S.C.R. 249, 2004 SCC 7

New Solutions Financial Corporation                                                              Appellant

v.

Transport North American Express Inc.                                                      Respondent

Indexed as:  Transport North American Express Inc. v. New Solutions Financial Corp.

Neutral citation:  2004 SCC 7.

File No.:  29355.

2003:  October 16; 2004:  February 12.

Present:  Iacobucci, Major, Bastarache, Arbour, LeBel, Deschamps and Fish JJ.

on appeal from the court of appeal for ontario

Contracts — Interest — Criminal rate — Illegality — Severability — Section 347  of Criminal Code  prohibiting effective annual interest rate in excess of 60 percent — Contract found to contravene prohibition — Offending interest rate read down so that contract provided for maximum legal rate of interest — Whether “notional severance” available as remedy in cases arising under s. 347 — Whether application of notional severance appropriate in this case — Criminal Code, R.S.C. 1985, c. C‑46, s. 347 .

The appellant and the respondent entered into a credit agreement pursuant to which the appellant advanced the respondent the sum of $500,000.  A commitment letter in respect of the proposed credit facility was signed and provided for the following payments: (a) interest at four percent per month calculated daily, payable monthly in arrears; (b) a monthly monitoring fee of $750; (c) a one percent standby fee; (d) royalty payments of $160,000 in eight quarterly installments; (e) payment of legal and other fees; and (f) a commitment fee of $5,000.  In addition to the commitment letter, the parties executed an accounts receivable factoring agreement, a promissory note, a general security agreement and personal guarantees of the indebtedness for up to $500,000.  From the outset, the parties had agreed to depart from the terms of the accounts receivable factoring agreement.  The application judge granted the respondent’s application for a declaration that the agreement contained an interest component that contravened s. 347  of the Criminal Code  and applied “notional severance” to reduce the effective annual interest rate to 60 percent so the agreement would comply with s. 347.  The majority of the Court of Appeal allowed the respondent’s appeal; it struck out the clause providing for interest at a rate of four percent per month, calculated daily and payable monthly in arrears, and left in place the other payments, which amounted to an effective annual rate of 30.8 percent when computed as interest as per s. 347(2).

Held (Bastarache, Deschamps and Fish JJ. dissenting):  The appeal should be allowed.

Per Iacobucci, Major, Arbour and LeBel JJ.:  The various payments made by the respondent, with the exception of any portion of the payments relating to repayment of principal, satisfy the definition of “interest” in s. 347(2) and cumulatively amount to an interest rate in excess of that permitted by the Code.  Notional severance is available as a matter of law as a remedy in cases arising under s. 347.  The traditional rule that contracts in violation of statutory enactments are void ab initio is not the approach courts should necessarily take in cases of statutory illegality involving s. 347.  Instead, judicial discretion should be employed and a spectrum of remedies is available.  At one end of the spectrum are contracts so objectionable that their illegality will taint the entire contract — exploitive loan-sharking arrangements and contracts that have a criminal object should be declared void ab initio.  At the other end of the spectrum are contracts that, although they do contravene a statutory enactment, are otherwise unobjectionable.  Contracts of this nature will often attract the application of the doctrine of severance.  In each case, the determination of where along the spectrum a given case lies, and the remedial consequences flowing therefrom, will hinge on a careful consideration of the specific contractual context and the illegality involved.  If the case is an appropriate one for the court to sever only those provisions of the loan agreement that put the effective interest rate over 60 percent, and if it is conceded, as it must be, that such rewording alters the agreement of the parties, the question becomes only a choice of the appropriate technique of severance.  The preferred severance technique is the one that, in light of the particular contractual context involved, would most appropriately cure the illegality while remaining otherwise as close as possible to the intentions of the parties expressed in the agreement.  The “blue-pencil” technique may not necessarily achieve that result.  The change effected by the blue-pencil technique will often fundamentally alter the consideration associated with the bargain and do violence to the intention of the parties.  Indeed, in many cases, the application of the blue-pencil approach will provide for an interest-free loan where the parties demonstrated in the agreement a clear intention to charge and pay considerable interest.

The application of notional severance to the agreement between the parties in this case was appropriate.  Four considerations are relevant to the determination of whether public policy ought to allow an otherwise illegal agreement to be partially enforced rather than being declared void ab initio in the face of illegality in the contract:  (1) whether the purpose or policy of s. 347 would be subverted by severance; (2) whether the parties entered into the agreement for an illegal purpose or with an evil intention; (3) the relative bargaining positions of the parties and their conduct in reaching the agreement; (4) the potential for the debtor to enjoy an unjustified windfall.  Given that the present case involved a commercial transaction engaged in by experienced and independently advised commercial parties, it is difficult to see why the choice of a 30.8 percent rather than 60 percent rate better fosters compliance with s. 347(1)(a) of the Code.  The other considerations also militate in favour of a flexible remedy.  There is no evidence on the record to suggest that the appellant has been charged with violating s. 347(1)(a).  The contract was entered into for ordinary commercial purposes and there was nothing inherently illegal or evil about this intention.  With respect to the third factor, each party in this case had independent legal advice, was commercially experienced and knew what it was getting into.  Finally, any potential for an unjustified windfall in this case arises from the respondent possibly not having to repay the principal and interest, or from the respondent possibly not having to pay a commercially appropriate rate of interest on the loan.  Given that each party had independent legal advice and knew precisely the obligations that it was taking on, the equities of the situation favour the appellant.

Per Deschamps and Fish JJ. (dissenting):  The well-established “blue-pencil” remedy applied in the Court of Appeal respects the trial judge’s findings of fact, as it must, and achieves an equitable result consistent with established principle.  The trial judge erred in straining the recognized rules of equity by resorting to what he described as “notional severance”.  Even on the assumption that notional severance is available as a matter of law as a remedy in cases arising under s. 347 of the Code, notional severance should be permitted only where:  (1) public policy does not require that the entire agreement be declared unenforceable; (2) severance is found to be warranted; and (3) severance simpliciter — or “blue-pencil severance” — is impracticable or would occasion an unjust result.  Here, on the facts as found by the trial judge, the first two criteria are satisfied but the third is not:  blue-pencil severance is both possible and fair.  Unlike notional severance as applied by the trial judge, blue-pencil severance does not do violence to the policy purposes of s. 347 of the Code or require a judicial re-writing of the interest clause agreed to, as such, by the parties.  It appears neither “artificial” nor “arbitrary” to sever the criminal rate of interest agreed to as interest and to leave intact the distinct and separate charges not agreed to as interest, and not considered to be interest by either party.  This solution still leaves the appellant with a return of slightly more than 30 percent per annum.  Further, it appears that all four relevant considerations, examined individually and weighed together, militate here in favour of the remedy applied by the Court of Appeal and against the remedy adopted by the trial judge.  The effect of the trial judge’s decision was to stretch the principles of equity in an inappropriate way and to send the wrong message to those who lend money at a criminally prohibited rate to “willing” borrowers who cannot otherwise obtain a loan. They should not be encouraged to believe that if their illegal arrangement is subjected to judicial scrutiny, they will nonetheless recover the highest rate they could legally have charged — and thus suffer no pecuniary disadvantage for having violated s. 347 of the Code.

Per Bastarache J. (dissenting): Severance in the traditional sense, not “notional severance”, should be the remedy available in these circumstances.  There is a fundamental difference between striking out offending sections of a contract and rewriting a central provision.  Although both approaches interfere in some way with the intent of the parties, the added flexibility of the rewriting approach comes at a considerable cost and it is not supported by any principle of contract law.  By contrast, the severance doctrine is a long-standing one.  The “blue-pencil” test has been applied several times in the context of violations of s. 347 of the Code to strike out offending interest provisions.  Moreover, there is no legal or other principled reason to limit the application of the new approach endorsed by the majority to the criminal rate of interest and other illegal provisions would be open to judicial redrafting.  The availability of “notional severance” as a remedy creates greater uncertainty in the law.  Under notional severance, courts will be permitted to literally add new words to the parties’ agreement and by doing so, courts will be substituting their intentions for those of the parties.

The approach taken by the majority in this case is also inconsistent with that taken in Garland v. Consumers’ Gas Co., [1998] 3 S.C.R. 112, where the Court interpreted the definition of interest broadly in order to prevent creditors from avoiding the statute by manipulating the form in which payments were made.  Under the majority’s approach, a creditor would be permitted to escape the consequences of its avoidance measures by simply reducing the rate applied to the maximum permitted under the Code and this approach is inconsistent with the general objectives expressed in the Code and incompatible with the notion of deterrence.  In addition, there do not appear to be compelling reasons to depart from well-established precedent in this case.

Lastly,  clause 9.1 of the accounts receivable factoring agreement reflects the parties’ own intentions as to remedy.  It cannot be said that the wording of the clause anticipates — still less, provides for — notional severance.  In view of the plain meaning of clause 9.1, counsel’s concession as to the intention of the parties, and the letter sent by the appellant’s solicitor to the respondent three weeks after the commitment letter was signed, the blue-pencil approach adopted by the Court of Appeal is entirely consistent with the parties’ own intentions and notional severance is not.

Cases Cited

By Arbour J.

Applied:  William E. Thomson Associates Inc. v. Carpenter (1989), 61 D.L.R. (4th) 1; referred to:  Garland v. Consumers’ Gas Co., [1998] 3 S.C.R. 112; Still v. M.N.R., [1998] 1 F.C. 549; Cope v. Rowlands (1836), 2 M. & W. 149, 150 E.R. 707; Kocotis v. D’Angelo (1957), 13 D.L.R. (2d) 69; Bank of Toronto v. Perkins (1883), 8 S.C.R. 603; Neider v. Carda of Peace River District Ltd., [1972] S.C.R. 678; Mira Design Co. v. Seascape Holdings Ltd., [1982] 4 W.W.R. 97; Trillium Computer Resources Inc. v. Taiwan Connection Inc. (1993), 11 B.L.R. (2d) 1, aff’d (1994), 11 B.L.R. (2d) 1; Milani v. Banks (1997), 145 D.L.R. (4th) 55.

By Fish J. (dissenting)

William E. Thomson Associates Inc. v. Carpenter (1989), 61 D.L.R. (4th) 1.

By Bastarache J. (dissenting)

Still v. M.N.R., [1998] 1 F.C. 549; Cope v. Rowlands (1836), 2 M. & W. 149, 150 E.R. 707; Bank of Toronto v. Perkins (1883), 8 S.C.R. 603; Steinberg v. Cohen, [1930] 2 D.L.R. 916; Hasiuk v. Oshanek, [1936] 1 D.L.R. 232; Carney v. Herbert, [1985] 1 All E.R. 438; McFarlane v. Daniell (1938), 38 S.R. 337; Attwood v. Lamont, [1920] 3 K.B. 571; Canadian American Financial Corp. (Canada) Ltd. v. King (1989), 60 D.L.R. (4th) 293; Mira Design Co. v. Seascape Holdings Ltd., [1982] 4 W.W.R. 97; Garland v. Consumers’ Gas Co., [1998] 3 S.C.R. 112; Friedmann Equity Developments Inc. v. Final Note Ltd., [2000] 1 S.C.R. 842, 2000 SCC 34.

Statutes and Regulations Cited

Criminal Code , R.S.C. 1985, c. C‑46 , s. 347 .

Authors Cited

Fridman, G. H. L.  The Law of Contract in Canada, 4th ed.  Scarborough, Ont.:  Carswell, 1999.

Marsh, Norman S.  “The Severance of Illegality in Contract” (1948), 64 L.Q.R. 230 and 347.

Waddams, S. M.  The Law of Contracts, 4th ed.  Toronto:  Canada Law Book, 1999.

Ziegel, Jacob S.  “Bill C‑44:  Repeal of the Small Loans Act and Enactment of a New Usury Law” (1981), 59 Can. Bar Rev. 188.

Ziegel, Jacob S.  “The Usury Provisions in the Criminal Code :  The Chickens Come Home to Roost” (1986), 11 Can. Bus. L.J. 233.

APPEAL from a judgment of the Ontario Court of Appeal (2002), 60 O.R. (3d) 97, 214 D.L.R. (4th) 44, 160 O.A.C. 381, 27 B.L.R. (3d) 163, 6 R.P.R. (4th) 1, [2002] O.J. No. 2335 (QL), reversing a judgment of the Superior Court of Justice (2001), 54 O.R. (3d) 144, 200 D.L.R. (4th) 560, 16 B.L.R. (3d) 148, [2001] O.J. No. 1948 (QL).  Appeal allowed, Bastarache, Deschamps and Fish JJ. dissenting.

Peter J. Cavanagh and Eric N. Hoffstein, for the appellant.

Robert G. Ackerman, for the respondent.

The judgment of Iacobucci, Major, Arbour and LeBel JJ. was delivered by

Arbour J. —

I.  Overview

1                                   In March 2000, the appellant, New Solutions Financial Corp. (“New Solutions”), and the respondent, Transport North American Express Inc. (“TNAE”), entered into a credit agreement pursuant to which New Solutions advanced TNAE the sum of $500,000.  In addition to various other fees and charges, the agreement provided for interest to be paid at the rate of four percent per month, calculated daily and payable monthly in arrears.  By all accounts, the various payments called for by the agreement constituted a “criminal rate” of interest as defined in s. 347  of the Criminal Code , R.S.C. 1985, c. C‑46  (the “Code ”).  The payments soon became too onerous for TNAE to meet, and the company applied to the Ontario Superior Court of Justice for a declaration that the agreement contained an illegally high rate of interest and should not be enforced.

2                                   The application judge, Cullity J., ruled that he was not confined to the so-called “blue-pencil” approach to severance in dealing with the statutory illegality of the contract, whereby only discrete illegal promises could be excised.  Using “notional severance”, he read down the offending interest rate so the contract provided for the maximum legal rate of interest:  (2001), 54 O.R. (3d) 144.

3                                   Upon appeal to the Court of Appeal for Ontario, Rosenberg J.A., for the majority, concluded that the doctrine of severance only permits the striking of distinct promises from a contract: (2002), 60 O.R. (3d) 97.  He reversed the application judge’s finding that notional severance was an available remedial instrument.  Rosenberg J.A. found that it was appropriate to strike out or blue-pencil the provision calling for interest at four percent per month, calculated daily and payable monthly in arrears, leaving the balance of the agreement to be enforced in accordance with its terms.  Sharpe J.A., agreeing with the reasons of Cullity J., dissented.

4                                   There is broad consensus that the traditional rule that contracts in violation of statutory enactments are void ab initio is not the approach courts should necessarily take in cases of statutory illegality involving s. 347  of the Code .  Instead, judicial discretion should be employed in cases in which s. 347 has been violated in order to provide remedies that are tailored to the contractual context involved.  The primary issue in this appeal by New Solutions is whether notional severance, as formulated and applied by Cullity J., is valid in Canadian law and applicable here.

5                                   Given the desirability of remedial flexibility in cases of statutory illegality arising in connection with s. 347  of the Code , the evolving nature of the law regarding statutory illegality generally and the sound policy basis  in which the concept is rooted, I find that notional severance is available as a matter of law as a remedy in cases arising under s. 347. 

6                                   A spectrum of remedies is available to judges in dealing with contracts that violate s. 347  of the Code .  The remedial discretion this spectrum affords is necessary  to cope with the various contexts in which s. 347 illegality can arise.  At one end of the spectrum are contracts so objectionable that their illegality will taint the entire contract.   For example, exploitive loan-sharking arrangements and contracts that have a criminal object should be declared void ab initio.  At the other end of the spectrum are contracts that, although they do contravene a statutory enactment, are otherwise unobjectionable.  Contracts of this nature will often attract the application of the doctrine of severance.  The agreement in this case is an example of such a contract.  In each case, the determination of where along the spectrum a given case lies, and the remedial consequences flowing therefrom, will hinge on a careful consideration of the specific contractual context and the illegality involved.

7                                   The application judge in this case found that (i) the agreement between New Solutions and TNAE only inadvertently violated s. 347; (ii) the parties were experienced in commercial matters and negotiated at arm’s length; (iii) there was no evidence that they did not have equal bargaining power; and (iv) they each had the benefit of independent legal advice in the course of the negotiations leading to the agreement.  Consequently, the application of notional severance to the agreement between New Solutions and TNAE in this case by Cullity J. was appropriate.  I would allow the appeal.

II.  Facts

8                                   For the relevant time period, TNAE was in the business of expedited freight trucking.  Ken and Karen Dragosits were shareholders in TNAE and actively involved in the operation of its business.  Prior to the end of 1999, other shareholders held a 50 percent interest in TNAE.  A corporation connected to these other shareholders provided TNAE the funds needed for the firm’s working capital.  A demand was made by this other corporation for the repayment of the funds owed to it by TNAE.  The Dragosits and TNAE decided to search out a source for the means to repay the indebtedness.

9                                   The Dragosits sought financing from BDO Capital, now the appellant, New Solutions, to enable TNAE to repay its indebtedness and for the other shareholders in TNAE to be bought out.  The parties eventually entered into an agreement that contained a high rate of interest and also significant other fees and charges.  The costly nature of the loan for TNAE no doubt reflected the high risk New Solutions was taking on in making the funds available.

10                               Before arriving at their agreement, New Solutions expressed interest in acquiring a 30 percent equity interest in TNAE in conjunction with the contemplated credit facility.  The Dragosits resisted this as they wished to be the sole shareholders of TNAE.  In lieu of surrendering an equity interest, they agreed that New Solutions would receive a “royalty payment” of $160,000, payable in eight quarterly installments, to reflect the approximate value of a 30 percent equity interest in TNAE.

11                               In the negotiations leading up to the agreement, each party had the benefit of independent legal advice.  On March 6, 2000, a commitment letter in respect of the proposed credit facility was signed by the Dragosits and provided for the following payments:

(a) interest at four percent per month calculated daily, payable monthly in arrears;

(b) a monthly monitoring fee of $750;

(c) a one percent standby fee;

(d) royalty payments of $160,000 in eight quarterly installments;

(e) payment of legal and other fees; and

(f) a commitment fee of $5,000.

With the exception of the standby fee, all these payments were found by Cullity J. and by Rosenberg J.A. to constitute “interest” under s. 347(2)  of the Code .  Presumably, the standby fee was not included in the calculation of the effective interest rate because no standby fees were charged since the full credit facility of $500,000 was drawn upon.

12                               By  March 30, 2000, the parties had, in addition to the commitment letter, executed an accounts receivable factoring agreement, a promissory note and a general security agreement.  The Dragosits also each executed personal guarantees of the indebtedness for up to $500,000 plus interest at the rate of 30 percent per annum.  From the outset, the parties had agreed to depart from the terms of the accounts receivable factoring agreement.  On March 28, 2000, the solicitor for New Solutions wrote to the solicitor for TNAE and the Dragosits, confirming that the parties had agreed on March 27 that they would not strictly follow the terms of the accounts receivable factoring agreement unless New Solutions elected to exercise its rights under it.  Instead, the understanding was that TNAE would borrow the full $500,000 from New Solutions and pay the interest, fees and royalties as set out in the commitment letter.  According to Cullity J., “the concept of a factoring of receivables was put aside and replaced by a revolving credit facility” (para. 5).

13                               The principal amount of $500,000 was advanced by New Solutions.    At the outset, TNAE paid interest at the rate of four percent per month, calculated daily and payable monthly in arrears, as well as the other fees and charges, in general accordance with the terms of the commitment letter.

14                               The various payments eventually became onerous, and TNAE sought legal advice regarding the repayment of the borrowed funds.  TNAE then applied to  the Ontario Superior Court of Justice for a declaration that the agreement contained an interest component that contravened s. 347  of the Code .  It also sought an order that interest previously paid be returned.

15                               On the basis of actuarial evidence, Cullity J. found that the effective interest rate on the loan, if it was repaid in full within two years, was 90.9 percent per annum.  In itself, the promise to pay interest at four percent per month calculated daily, payable monthly in arrears, amounted to an effective annual interest rate of 60.1 percent.  The remaining payments amounted to an effective annual interest rate of  30.8 percent.

16                               New Solutions originally denied that the agreement violated the Code  but sought severance and rectification if it did.  Cullity J. found that the agreement was in contravention of s. 347(1)(a) and applied “notional severance” to reduce the effective annual interest rate to 60 percent so the agreement would comply with s. 347.  The Court of Appeal allowed TNAE’s appeal; it struck out  the clause providing for interest at a rate of four percent per month calculated daily and payable monthly in arrears, and left in place the other payments, which amounted to an effective annual rate of 30.8 percent when computed as interest as per s. 347(2).  New Solutions seeks the restoration of the decision of the application judge.

III.  Relevant Statutory Provisions

17                               The pertinent text of the relevant provision of the Criminal Code  is:

347. (1) Notwithstanding any Act of Parliament, every one who

(a) enters into an agreement or arrangement to receive interest at a criminal rate, or

(b) receives a payment or partial payment of interest at a criminal rate,

is guilty of

(c) an indictable offence and is liable to imprisonment for a term not exceeding five years, or

(d) an offence punishable on summary conviction and is liable to a fine not exceeding twenty-five thousand dollars or to imprisonment for a term not exceeding six months or to both.

(2) In this section,

“credit advanced” means the aggregate of the money and the monetary value of any goods, services or benefits actually advanced or to be advanced under an agreement or arrangement minus the aggregate of any required deposit balance and any fee, fine, penalty, commission and other similar charge or expense directly or indirectly incurred under the original or any collateral agreement or arrangement;

“criminal rate” means an effective annual rate of interest calculated in accordance with generally accepted actuarial practices and principles that exceeds sixty per cent on the credit advanced under an agreement or arrangement;

                                                                   . . .

“interest” means the aggregate of all charges and expenses, whether in the form of a fee, fine, penalty, commission or other similar charge or expense or in any other form, paid or payable for the advancing of credit under an agreement or arrangement, by or on behalf of the person to whom the credit is or is to be advanced, irrespective of the person to whom any such charges and expenses are or are to be paid or payable, but does not include any repayment of credit advanced or any insurance charge, official fee, overdraft charge, required deposit balance or, in the case of a mortgage transaction, any amount required to be paid on account of property taxes;

                                                                   . . .

(3) Where a person receives a payment or partial payment of interest at a criminal rate, he shall, in the absence of evidence to the contrary, be deemed to have knowledge of the nature of the payment and that it was received at a criminal rate.

                                                                   . . .

(7) No proceedings shall be commenced under this section without the consent of the Attorney General.

IV.  Issue

18                               Are judges in Canada permitted by law to exercise remedial discretion to partially enforce a contract contravening s. 347  of the Code  by reading down interest rate provisions to avoid what would otherwise be illegality?

V.  Analysis

A.  Illegality of the Contract

19                               The definition of “interest” in s. 347(2) is broad: see Garland v. Consumers’ Gas Co., [1998] 3 S.C.R. 112, at para. 28.  The various payments made by TNAE, with the exception of any portion of the payments relating to the repayment of principal,  satisfy the definition of “interest” as defined in  s. 347(2).  This includes the “royalty payments”.  I  agree with the courts below that the payments made by TNAE to New Solutions cumulatively amount to an interest rate in excess of that permitted by the Code .

B.  The Doctrine of Illegality

20                               The Federal Court of Appeal’s decision in Still v. M.N.R., [1998] 1 F.C. 549, provides a useful summary of the development of the doctrine of illegality, including a discussion of the development and evolution of the doctrine’s common law and statutory branches.  In addressing the current state of the doctrine of illegality, Robertson J.A. remarked, at para. 12:

Law reform agencies have been quick to conclude that the law of illegality is in an unsatisfactory state . . . . There is a plethora of conflicting decisions and great uncertainty as to the principles which should be guiding the courts.  Arguably, so many exceptions have been grafted on to the common law rule that illegal contracts are void ab initio that the validity of the rule itself is brought into question.

In light of the excellent treatment of the doctrine’s history by Robertson J.A. in Still v. M.N.R., there would be little benefit to fully retracing the doctrine’s history here.  Instead, given the evolving nature of this area of law, a very brief survey of some of the existing case law on the application of the doctrine of illegality will provide sufficient context for the finding in this case that notional severance is available as a discretionary remedy in cases where s. 347 has been violated.

21                               The  historical common law approach to contractual illegality is reflected in the following passage of Parke B. in Cope v. Rowlands (1836), 2 M. & W. 149, 150 E.R. 707 (Ex. Ct.), at p. 710:

[W]here the contract which the plaintiff seeks to enforce, be it express or implied, is expressly or by implication forbidden by the common or statute law, no court will lend its assistance to give it effect.  It is equally clear that a contract is void if prohibited by a statute, though the statute inflicts a penalty only, because such a penalty implies a prohibition.

In Cope v. Rowlands, the question surrounded whether an unlicensed broker could recover for the work that he had done for the defendant.  The court concluded that the legal requirement (under threat of penalty) that brokers be licensed by the city of London implied a prohibition on work being done by unlicenced brokers.  As a consequence, the contract was held to be void ab initio and the unlicensed broker was unable to enforce his claim for payment for the work that had been done.  The Court of Appeal for Ontario denied recovery in a similar case involving an electrician seeking to recover for work done without possessing the appropriate class of licence:  see Kocotis v. D’Angelo (1957), 13 D.L.R. (2d) 69. 

22                               The historical common law approach that contracts illegal under statute are void ab initio has been applied by this Court: see, e.g., Bank of Toronto v. Perkins (1883), 8 S.C.R. 603, and more recently, Neider v. Carda of Peace River District Ltd., [1972] S.C.R. 678.  However, some time ago Canadian courts began to develop a more flexible approach to statutory illegality in contract, often severing the illegal provisions and enforcing the remainder.  For example, in one of the earliest cases dealing with the application of s. 347  of the Code , Mira Design Co. v. Seascape Holdings Ltd., [1982] 4 W.W.R. 97 (B.C.S.C.), Huddart L.J.S.C. held that although the interest provisions of a mortgage were unenforceable, exceeding as they did the maximum effective interest rate permitted under s. 305.1 of the Code  (the predecessor to s. 347), the contract as a whole should not be held to be void ab initio.  Her reasoning, at p. 104, was that although the section makes it an offence to receive interest at an illegal rate, the section did not seek to make associated collateral agreements (such as for the transfer of the real estate or the payment of the principal amount owing on the mortgage) void ab initio:

Most Canadians would agree that the purpose of the Criminal Code  is to protect the public by providing for the punishment of behaviour that Parliament considers to be against the public interest.  The purpose of s. 305.1 [now s. 347] is to punish everyone who enters into an agreement or arrangement to receive interest at a criminal rate.  It does not expressly prohibit such behaviour, nor does it declare such an agreement or arrangement to be void.  The penalty is severe, and designed to deter persons from making such agreements.  It replaces the Small Loans Act, which included a prohibition of such agreements and gave the court the power to reconstruct them.  It is designed to protect borrowers.  There is no penalty imposed on a person who makes an agreement to pay, or pays, interest at a criminal rate.  It is not designed to prevent persons from entering into lending transactions per se.

23                               The same approach was taken by the Court of Appeal for Ontario in William E. Thomson Associates Inc. v. Carpenter (1989), 61 D.L.R. (4th) 1.  Having considered s. 347  of the Code , the court in that case concluded that where an interest rate provided for in an agreement exceeds the 60 percent statutory maximum, the interest rate provision of the contract may be severed without declaring the whole contract void.

24                               In Thomson, at p. 8, Blair J.A. considered the following four factors in deciding between partial enforcement and declaring a contract void ab initio: (i) whether the purpose or the policy of s. 347 would be subverted by severance; (ii) whether the parties entered into the agreement for an illegal purpose or with an evil intention; (iii) the relative bargaining positions of the parties and their conduct in reaching the agreement; and (iv) whether the debtor would be given an unjustified windfall.  He did not foreclose the possibility of applying other considerations in other cases, however, and remarked (at p. 12) that whether “a contract tainted by illegality is completely unenforceable depends upon all the circumstances surrounding the contract and the balancing of the considerations discussed above and, in appropriate cases, other considerations”.

25                               In Trillium Computer Resources Inc. v. Taiwan Connection Inc. (1993), 11 B.L.R. (2d) 1 (Ont. Ct. (Gen. Div.)), aff’d (1994), 11 B.L.R. (2d) 1 (Ont. Div. Ct.), Conant J. entered summary judgment in favour of the plaintiff who had paid $8,000 interest in consideration of credit extended by the defendant for eight days.  In a brief judgment and without addressing the authorities on this point, Conant J. stated, at p. 2:

I am satisfied that an interest rate of over 3,000 % per annum, whether it be for credit and/or compensation for damages and other matters suffered by the Defendant, is a flagrant breach of s. 347  of the Criminal Code  of Canada .  This, in my view, is illegal and shall be returned to the Plaintiff less the maximum rate of 60% per annum allowed under the Code .  [Emphasis added.]

This approach is similar to the one applied by Cullity J. and endorsed by Sharpe J.A. (in dissent at the Court of Appeal) in the present case.

26                               In Milani v. Banks (1997), 145 D.L.R. (4th) 55, the Court of Appeal for Ontario applied the contextual approach endorsed by Blair J.A. in Thomson, supra.  This case involved a $32,000 loan with a term of 30 days.  The contract provided for $3,000 to be kept by the creditor in respect of the costs associated with the loan, and an 18 percent annualized interest rate to be paid on the full principal amount.  McKinlay J.A. for the court held, at pp. 59‑60 that:

In this case, the appellant takes the position that the only offensive part of the loan was the $3,000 charge for “fees”, and that if the agreement were left intact apart from that provision, the result would be a fair one in the circumstances.  I am inclined to agree with that position . . . .

. . .

I consider this case to be one strongly favouring the position of the appellant.  She is clearly not entitled to the $3,000 fee, but I would strike only that provision, and leave the loan otherwise intact as a $32,000 loan with interest at 18% per annum for a thirty day term.

The approach taken by McKinlay J.A. in Milani is reflected in the path taken by Rosenberg J.A. at the Court of Appeal for Ontario in the case at bar.  McKinlay J.A. severed one of the “interest” terms (actually attributable to “costs”) from the loan so that the interest rate would be legal, just as Rosenberg J.A. in this case severed the promise to pay interest at four percent per month, calculated daily, payable monthly in arrears, thereby leaving the other charges to amount, cumulatively, to a permissible rate of interest under s. 347.

C.   The Problematic Nature of the Blue-Pencil Test

27                               The blue-pencil approach is understood both as a test of the availability of severance to remedy contractual illegality and also as a technique for effecting severance.  The blue-pencil approach as a test of the appropriateness of severance requires a consideration of whether an illegal contract can be rendered legal by striking out (i.e., by drawing a line through) the illegal promises in the agreement.  The resulting set of legal terms should retain the core of the agreement.  If the nature or core of the agreement is disturbed, then on this test the illegal clause in the contract is not a candidate for severance and the entire contract is void.  The blue-pencil approach as a technique of effecting severance involves the actual excision of the provisions leading to the illegality, leaving those promises untainted by the illegality to be enforced.

28                               The use of the blue-pencil approach to sever one or more provisions from a contract alters the terms of the agreement between the parties.  The only agreement that one can say with certainty the parties would have agreed to is the one that they actually entered into.  The insistence in the case law that the blue-pencil test derives its validity from refusing to change or add words or provisions to the contract is unconvincing.  It is doubtful, for example, that the lenders in cases such as Thomson, supra, or Mira Design, supra, would have entered into the agreements at issue had they been aware ex ante that they would only be entitled to the return of the principal advanced.  The change effected by the blue-pencil technique will often fundamentally alter the consideration associated with the bargain and do violence to the intention of the parties.  Indeed, in many cases, the application of the blue-pencil approach will provide for an interest-free loan where the parties demonstrated in the agreement a clear intention to charge and pay considerable interest.

29                               The blue-pencil test was developed in cases where the courts were considering instruments under seal, where the form of the deed governed and where the intention of the parties was  irrelevant.  It was therefore important that what remained after severance would be a valid deed:

In the deed form was everything; the actual intention of the parties was immaterial.  It was, therefore, natural that in considering the possibility of severance of promises in a deed, the court should be concerned to see that what was left remained a valid deed; there could be no question of implying a promise to take effect if part of the original bargain was illegal.  This is the historical origin of what was later called the ‘blue-pencil test’.

(N. S. Marsh, “The Severance of Illegality in Contract” (1948), 64 L.Q.R. 230 and 347, at pp. 351‑52)

Historically, courts were not concerned with the intention of the parties.  The artificiality of the blue-pencil test arises from the common law constraints imposed on courts unaided by principles of equity.

30                                Courts inescapably make a new bargain for the parties when they use the blue-pencil approach.  As Cullity J. remarked, at paras. 35‑36:

The blue-pencil test is, I believe, a relic of a bygone era when the attitude of courts of common law — unassisted by principles of equity — towards the interpretation and enforcement of contracts was more rigid than is the case at the present time.  At an early stage in the development of the law relating to illegal promises, severance was held to be justified on the basis of the blue-pencil test alone.  As the reasoning in Milani and William E. Thomson demonstrates, we have moved a long way beyond that mechanical approach.  Enforcement may be refused in the exercise of the kind of discretionary judgment I have mentioned even where blue-pencil severance is possible.

Despite repeated statements in the cases that the court will not make a new agreement for the parties, that is, of course, exactly what it does whenever severance is permitted in cases like William E. Thomson and Milani.  [Emphasis added.]

I am in complete agreement with the conclusion that when a court employs the blue-pencil test, it is making a new agreement for the parties. Indeed, all forms of severance alter the terms of the original agreement.

31                                I also agree with the view of Rosenberg J.A. at the Court of Appeal in the present case, at para. 33, that severance lies along a spectrum of available remedies.  Depending on the circumstances, the court may exercise its discretion to find the whole agreement unenforceable or sever only the provision(s) that put the effective interest rate over 60 percent:

[A] judge has discretion to apply the doctrine of severance to an agreement that offends the criminal interest rate provisions of the Code . This discretion gives rise to a spectrum of available remedies. Where the loan transaction resembles a traditional loan sharking arrangement, the court may refuse to apply the doctrine of severance and hold the entire loan agreement unenforceable, including the obligation to repay the principal. While this remedy leaves the borrower with a windfall, this result may be justified in some cases by the need to denounce such usurious practices. See C.A.P.S. International Inc. v. Kotello, [2002] M.J. No. 205 [(QL)] (Q.B.). At the other end of the spectrum, in the case of a good faith commercial transaction where the equities favour the lender and severance does not undermine the policy of the legislation, the court may sever only those provisions of the loan agreement that put the effective interest rate over 60 per cent, leaving intact the borrower’s obligation to repay the principal and pay some interest. See e.g., Milani, supra. Closer to the centre of the spectrum lies a case like Terracan [Capital Corp. v. Pine Projects Ltd. (1993), 100 D.L.R. (4th) 431(B.C.C.A.)], where the court severed all the interest provisions but upheld the debtor’s obligation to repay the principal.  [Emphasis added.]

This statement of the remedial discretion of a judge in a case involving a violation of s. 347  of the Code   takes into account the seriousness of the illegality involved in any given case, the identity and nature of the parties and the broader contractual context.

32                               If the case is an appropriate one for the court to sever only those provisions of the loan agreement that put the effective interest rate over 60 percent, and if it is conceded, as it must be, that such a rewording alters the agreement of the parties, the question becomes only a choice of the appropriate technique of severance.  The preferred severance technique is the one that, in light of the particular contractual context involved, would most appropriately cure the illegality while remaining otherwise as close as possible to the intentions of the parties expressed in the agreement.  The blue-pencil technique may not necessarily achieve that result.

33                               The blue-pencil test is imperfect because it involves mechanically removing illegal provisions from a contract, the effects of which are apt to be somewhat arbitrary.  The results may be arbitrary in the sense that they will be dependent upon accidents of drafting and the form of expression of the agreement, rather than the substance of the bargain or consideration involved.  For example, if the effective interest rate of the total interest obligation (as defined in s. 347(2)) in the agreement between New Solutions and TNAE were only 0.1 percent lower, then the excision of the various charges, fees and royalty payment provisions using the blue-pencil technique would have resulted in a legally valid agreement bearing an effective annual rate of interest of 60 percent.  Although the results obtained from the blue-pencil approach will in many cases be sensible and may often be desirable, due to its artificiality, the application of the blue-pencil approach will sometimes be inappropriate.

34                                Section 347  of the Code  invites difficulties with arbitrariness by imposing a bright line of 60 percent as demarcating legal interest from illegal interest.  This legislatively mandated bright line distinguishes s. 347 cases from those involving provisions, for example, in restraint of trade, where there is no bright line.  The interaction of the blue-pencil test with the bright line separating illegal interest from legal interest leads to erratic results.  Consider the following three substantively equivalent contracts.  Assume for each contract that each party was commercially sophisticated, of equivalent bargaining power and in receipt of independent legal advice.  Assume also that neither party was aware the interest payable was prohibited by the Code .  The three contracts lead to dramatically different results when the blue-pencil test is applied.

35                               First, consider the result obtained using the blue-pencil approach in the case at bar.  Accepted actuarial evidence showed that the contract between TNAE and New Solutions providing for interest at four percent per month, calculated daily, payable monthly in arrears, represented an effective interest rate of 60.1 percent per annum; other fees and charges reflected a further 30.8 percent interest per annum, resulting in a total rate of 90.9 percent.  The application of the blue-pencil test required that, at the very least, the 60.1 percent interest provision be struck from the contract since, standing alone, it violated s. 347.  After considering the equities of the contractual context, Rosenberg J.A. concluded that TNAE should have to repay the principal and at least some interest.  Consequently, the provision calling for 60.1 percent effective annual interest was struck from the contract, and TNAE was held accountable for repaying the principal plus the other fees and charges, amounting to an effective annual interest rate of 30.8 percent.

36                               Second, consider a contract that provides for 60.0 percent interest in one provision and 30.9 percent interest in the remaining provisions.  This contract would result in the same 90.9 percent per annum rate of interest payable overall.  Given the equities of the contractual context as found in the courts below, the application of the blue-pencil test would result in the other fees and charges being struck from the contract and the 60.0 percent interest payment obligation surviving.

37                               Finally, consider what the result would be under the blue-pencil approach  if there were only one interest provision, and that this provision called for an effective annual interest rate of 90.9 percent, which would also involve exactly the same amount of consideration passing from borrower to the lender as under the previous two contracts.  The blue-pencil test would require that the entire interest obligation be severed from the contract, since it alone would violate s. 347.  This would result in no interest being payable under the contract.  It is possible, in this last example, that the illegality of the rate would be so apparent that courts would be justified in reducing to zero such a blatantly illegal rate.  On the other hand, if the criminal rate was arrived at inadvertently — or under a misapprehension of the law — as assumed for the comparison of these three contracts, the obligation to repay at least some interest should be upheld.  In that case, the maximum permissible rate would best reflect the intention of the parties, while curing the illegality.

38                               This demonstrates that, at least in the case of contracts in violation of s. 347  of the Code , the results associated with the application of the blue-pencil approach are overly dependent upon the form of the contract, rather than its substance.  In these three similar contracts, the borrower must pay 30.8 percent interest, 60.0 percent interest, or no interest, depending on otherwise insignificant differences in drafting.  Results this erratic and sensitive  to the form of contractual expression are undesirable, and can be avoided through the use of notional severance in cases where considerations flowing from the broad contractual context favour the lender.

39                               There is little danger that abuses of this remedial flexibility would surreptitiously creep past trial judges. On the contrary, judges are apt to be quite suspicious, and rightly so, of credit arrangements which provide for effective annual rates of interest in excess of 60 percent per annum.  Using notional severance to read down interest provisions to be just within the legal limit would not find application in traditional loan-sharking transactions.  It would be available as a remedy where a court recognizes the commercial sophistication and professional advice received by both parties, concludes that the violation of s. 347 by the parties was unintentional, and considers it equitable to give effect to the highest legal interest obligation available. In such cases, there is no reason, in my view, to prefer the blue-pencil approach as more likely to deter criminal interest rates.  I will return to this point below.

40                               Thus, the appropriate approach is to vest the greatest possible amount of remedial discretion in judges in courts of first instance.  The spectrum of available remedies runs from a court holding contracts in violation of s. 347 void ab initio, in the most egregious and abusive cases, according to the criteria identified in Thomson, supra, to notional severance.  In the determination of where along the spectrum a particular contract lies, the considerations identified in Thomson by Blair J.A. should be referred to and analysed carefully.  Although Blair J.A. was considering the desirability of severing illegal interest from principal, the same factors are helpful in determining whether to reduce illegal interest to a legal level.

D.   Application of the Revised Approach

41                               Finding, as I do, that the maximum level of remedial flexibility should be vested in judges and available for application by them subject to a careful analysis of the factors identified in Thomson, I will apply this approach to the facts of the case at bar.

42                               As outlined above, in Thomson, Blair J.A. identified four considerations relevant to the determination of whether public policy ought to allow an otherwise illegal agreement to be partially enforced rather than being declared void ab initio in the face of illegality in the contract:

1.  whether the purpose or policy of s. 347 would be subverted by severance;

2.  whether the parties entered into the agreement for an illegal purpose or with an evil intention;

3.  the relative bargaining positions of the parties and their conduct in reaching the agreement;

4.    the potential for the debtor to enjoy an unjustified windfall.

43                               The first  factor — whether the policy behind s. 347 would be subverted here by partial enforcement — is related to the concerns expressed by Rosenberg J.A. in this case that civil remedies should not frustrate the deterrence purpose of the criminal prohibition.  In this regard, it is important to identify the policy purpose underlying s. 347.  Ostensibly, it was intended to curb loan-sharking activity.  It is difficult to pinpoint the precise rationale behind the provision, however, because the legislative record yields few clues.  According to Professor Ziegel:

On July 22nd, 1980, only a day after its first reading, the House of Commons gave second and third reading to Bill C‑44, “An Act to amend the Small Loans Act and to provide for its repeal and to amend the Criminal Code ”.  The Bill was not debated and its adoption had the unanimous support of all three political parties.  Any public discussion of the merits of the Bill was effectively forestalled by the haste with which it was rushed through the House.  This failure to give interested parties an opportunity to study and comment on the Bill would be serious enough if the Bill only dealt with minor technical matters.  But it does not.  The Bill deals with questions of major social, economic, and legal importance which warranted careful examination.  Even if one accepts the soundness of the objectives of the Bill, it does not follow that its technical implementation is equally unobjectionable or that the same goals could not have been realized in a less controversial manner.  In the writer’s view, the Bill is open to objections on both counts and may generate as many new problems as it was designed to resolve.

(J. S. Ziegel, “Bill C‑44: Repeal of the Small Loans Act and Enactment of a New Usury Law” (1981), 59 Can. Bar Rev. 188, at p. 188)

As was stated by this Court in Garland, supra, at para. 25:

The ostensible purpose of s. 347 was to aid in the prosecution of loan sharks.  See House of Commons Debates, 1st Sess., 32nd Parl., vol. III, July 21, 1980, at p. 3146; Thomson, supra, at p. 549.  However, it is clear from the language of the statute — e.g., its reference to insurance and overdraft charges, official fees, and property taxes in mortgage transactions — that s. 347 was designed to have a much wider reach, and in fact the section has most often been applied to commercial transactions which bear no relation to traditional loan-sharking arrangements.  Although s. 347 is a criminal provision, the great majority of cases in which it arises are not criminal prosecutions.  Rather, like the case at bar, they are civil actions in which a borrower has asserted the common-law doctrine of illegality in an effort to avoid or recover an interest payment, or to render an agreement unenforceable.  For this reason, the provision has attracted criticism from some commercial lawyers and academics, and calls have repeatedly been made for its amendment or repeal. . . .  Nevertheless, it is now well settled that s. 347 applies to a very broad range of commercial and consumer transactions involving the advancement of credit, including secured and unsecured loans, mortgages and commercial financing agreements.

Since it is very difficult to identify the policy objective behind s. 347  of the Code  beyond the prevention of loan-sharking, violations of the section that clearly do not involve loan-sharking should be approached cautiously, keeping in mind that there is no need to deter, through the criminal law, effective interest rates of up to 60 percent per year.  Given that this was a commercial transaction engaged in by experienced and independently advised commercial parties, it is difficult to see why the choice of a 30.8 percent rather than 60 percent rate better fosters compliance with s. 347(1) (a) of the Code .

44                               The second  factor is whether the agreement was entered into for an illegal purpose or with an evil intention.    There is no evidence on the record to suggest that New Solutions has been charged with violating s. 347(1)(a).  Absent a criminal conviction, New Solutions has the benefit of the presumption of innocence.   Concerns with specific and general deterrence are best addressed by the criminal law.  A prosecution under s. 347  of the Code  cannot be initiated without the consent of the Attorney General. This suggests that even a criminal remedy is not always appropriate for an infringement of s. 347, let alone a civil remedy seeking to promote the criminal law objective of deterrence.  Cullity J. found that the parties were unaware that the agreement contravened s. 347, their intent being only to provide the means by which TNAE could buy out its remaining shareholders and acquire commercially necessary working capital.   It was a contract  entered into for ordinary commercial purposes.  There was nothing inherently illegal or evil about this intention.  The worst that could be said is that TNAE was a high-risk debtor for New Solutions to lend to (even with the personal guarantees of the Dragosits), the corporation’s needs were high relative to its value, and it needed the money on relatively short notice.  This second consideration militates in favour of a flexible remedy.

45                               The third  factor concerns the relative bargaining positions of the parties and their conduct in reaching the agreement.  Each party had independent legal advice.  Each party was commercially experienced.  The Dragosits and TNAE knew what they were getting into, as did New Solutions.  The one failing seems to be that neither side realized that their agreement would contravene s. 347  of the Code .  This third consideration, too,  favours a flexible remedy.

46                               Finally, any potential for an unjustified windfall in this case arises from TNAE possibly not having to repay the principal and interest, or from TNAE possibly not having to pay a commercially appropriate rate of interest on the loan.  Given that each party had independent advice and knew precisely the obligations that they were taking on, my conclusion is that the equities of the situation favour New Solutions.  This is in accord with the findings by both the lower courts in this case.

47                               New Solutions should be repaid the principal, with the highest amount of interest legally allowable, namely 60 percent.  This represents a reduction from the interest rate of 90.9 percent that the contract provided for, but, given s. 347  of the Code , is the most that New Solutions can legally receive.  This no more rewrites the contract than enforcing only the other obligations, which cumulatively amount to “interest” at 30.8 percent.

48                               My colleague, Bastarache J., refers to clause 9.1 of the accounts receivable factoring agreement (para. 68), arguing it supports the blue-pencil approach to severance.  I agree that by clause 9.1 the parties expressed a preference for the use of severance to give full effect to the valid provisions of that agreement.  I have two responses to the view, however, that the clause calls for the use of the blue-pencil approach to severance rather than notional severance.  First, from the outset the parties agreed not to strictly abide by the terms of the accounts receivable factoring agreement.  Therefore, I question whether it is appropriate to  refer to clause 9.1 of this agreement.  Second, accepting for the sake of argument that the factoring agreement is relevant, the use of notional severance is not foreclosed by the intent underlying the inclusion of clause 9.1 in the factoring agreement.  The genuine intent underlying clause 9.1 was to preserve as much of the agreement as possible should any part of it be deemed to be invalid.  In light of the analysis above, the part of the agreement that is invalid is the obligation to pay an effective interest rate in excess of 60 percent per annum.  Therefore,  notional severance is  the most appropriate remedy to apply here, since it gives the greatest possible legal effect to the valid aspects of the agreement, which is in concert with the intention underlying clause 9.1.

VI.    Conclusion

49                               For the foregoing reasons I would allow the appeal, set aside the order of the Court of Appeal and restore the order of Cullity J.  I would award costs to the appellant in this Court and in the Court of Appeal.

The following are the reasons delivered by

50                               Bastarache J. (dissenting)  — I have read the reasons of Arbour J. and respectfully disagree.  I would not make available the remedy of “notional severance”, whereby judges are permitted to rewrite offending provisions in a contract so as to make them conform with the law.  Instead, I would restrict the remedy available under such circumstances to severance, in the traditional sense, whereby judges are permitted to strike out offending provisions in a contract provided they represent separate promises.

51                               My dissent is based on three major considerations.

52                               First, like Rosenberg J.A. of the Ontario Court of Appeal, I think there is a fundamental difference between striking out offending sections of a contract and rewriting a central provision of a contract.  Although both approaches admittedly interfere in some way with the intent of the parties, the added flexibility of the rewriting approach comes at a considerable cost.  Furthermore, it is not supported by any principle of contract law.  On the contrary, the severance doctrine is a long-standing one.  A useful survey of the history and development of severance for illegality can be found in Still v. M.N.R., [1998] 1 F.C. 549 (C.A.).

53                               Historically, courts held that illegality in contract rendered the entire agreement unenforceable and refused to order the return of money or property transferred under the agreement.  This classic approach to contractual illegality was articulated by Parke B. in Cope v. Rowlands (1836), 2 M. & W. 149, 150 E.R. 707 (Ex. Ct.), at p. 710: 

[W]here the contract which the plaintiff seeks to enforce, be it express or implied, is expressly or by implication forbidden by the common or statute law, no court will lend its assistance to give it effect.  It is equally clear that a contract is void if prohibited by a statute, though the statute inflicts a penalty only, because such a penalty implies a prohibition.

54                               This Court has also endorsed the traditional doctrine of statutory illegality. In Bank of Toronto v. Perkins (1883), 8 S.C.R. 603, Ritchie C.J. stated, at p. 610:

It would be a curious state of the law if, after the Legislature had prohibited a transaction, parties could enter into it, and, in defiance of the law, compel courts to enforce and give effect to their illegal transactions.

At p. 613, Strong J. added:

Whenever the doing of any act is expressly forbidden by statute, whether on grounds of public policy or otherwise, the English courts hold the act, if done, to be void, though no express words of avoidance are contained in the enactment itself.

See also Steinberg v. Cohen, [1930] 2 D.L.R. 916 (Ont. C.A.); Hasiuk v. Oshanek, [1936] 1 D.L.R. 232 (Man. C.A.).  

55                               Over time, the lower courts in Canada moved away from this strict approach because it was viewed as harsh and inequitable in some cases and could result in a windfall to one party.  Courts have come to apply a modern approach when faced with a contract that contains an illegal provision.  Under the modern approach to illegality, a contract that violates a statute may be enforceable, but not in its entirety.  In Carney v. Herbert, [1985] 1 All E.R. 438, the Privy Council succinctly summarized the principles to be applied.  At p. 443 of the decision, their Lordships quoted with approval the following statements made in an Australian case, McFarlane v. Daniell (1938), 38 S.R. 337 (N.S.W. Dist. Ct. App.), at p. 345:

When valid promises supported by legal consideration are associated with, but separate in form from, invalid promises, the test of whether they are severable is whether they are in substance so connected with the others as to form an indivisible whole which cannot be taken to pieces without altering its nature . . . .  If the elimination of the invalid promises changes the extent only but not the kind of the contract, the valid promises are severable . . . . If the substantial promises were all illegal or void, merely ancillary promises would be inseverable.

Accordingly, the illegal promise may be severed from the rest of the contract, leaving the remaining promises to be enforced by the court.  See G. H. L. Fridman, The Law of Contract in Canada (4th ed. 1999), at pp. 441‑45; S. M. Waddams, The Law of Contracts (4th ed. 1999), at p. 421.  This has been referred to as the “blue-pencil” test, a label first used in the case of Attwood v. Lamont, [1920] 3 K.B. 571 (C.A.).

56                               In my opinion, the “blue-pencil” test is nothing more than an expression to describe the application of the severance principle.  In Canadian American Financial Corp. (Canada) Ltd. v. King (1989), 60 D.L.R. (4th) 293 (B.C.C.A.), at pp. 299‑300, Hinkson J.A. referred to Attwood v. Lamont in which Lord Sterndale M.R. captured the essence of the applicable principle, at pp. 577‑78:

I think, therefore, that it is still the law that a contract can be severed if the severed parts are independent of one another and can be severed without the severance affecting the meaning of the part remaining.

This is sometimes expressed, as in this case by the Divisional Court, by saying that the severance can be effected when the part severed can be removed by running a blue pencil through it.

57                               Under the blue-pencil test, severance is only possible if the judge can strike out, by drawing a line through, the portion of the contract they want to remove, leaving the portions that are not tainted by illegality, without affecting the meaning of the part remaining.  In other words, the offending provision must constitute a separate promise, and one that is not part of the main purport and substance of the contract.  The provision must appear severable. See Fridman, supra, at p. 443.

58                               The blue-pencil test has been applied several times in the context of violations of s. 347  of the Criminal Code , R.S.C. 1985, c. C‑46 , to strike out offending interest provisions.  For example, in an earlier dealing with the application of s. 347 (then s. 305.1), Mira Design Co. v. Seascape Holdings Ltd., [1982] 4 W.W.R. 97 (B.C.S.C.), at p. 105, Huddart L.J.S.C. held:

In these circumstances, and considering that the offending provisions can be deleted by striking the first proviso on p. 2 of the mortgage, the words “in the event it is paid to the Mortgagee”, and “together with interest at the aforesaid rate on the principal sum of Eighty Thousand ($80,000) Dollars” from the second proviso, the entire following paragraph and the words “and interest” from the covenant contained in para. A on the same page, I have concluded that the offending portions are severable in fact.

At p. 104, Huddart L.J.S.C. also described the scope and application of s. 305.1 of the Criminal Code  as follows:

Most Canadians would agree that the purpose of the Criminal Code  is to protect the public by providing for the punishment of behaviour that Parliament considers to be against the public interest.  The purpose of s. 305.1 [now s. 347] is to punish everyone who enters into an agreement or arrangement to receive interest at a criminal rate.  It does not expressly prohibit such behaviour, nor does it declare such an agreement or arrangement to be void.  The penalty is severe, and designed to deter persons from making such agreements.  It replaces the Small Loans Act, which included a prohibition of such agreements and gave the court the power to reconstruct them.  It is designed to protect borrowers.  There is no penalty imposed on a person who makes an agreement to pay, or pays, interest at a criminal rate.  It is not designed to prevent persons from entering into lending transactions per se.  As regards subject matter, its scope and application are limited to agreements regarding interest.  Moreover, to find that s. 305.1 necessarily prohibits the entering into of agreements or arrangements to receive interest at a criminal rate would be to accomplish that which Parliament has chosen not to do, or cannot do, directly.

59                               Second, there is no legal or other principled reason to limit the application of the new approach endorsed by my colleague, that is notional severance, to the application of the criminal rate of interest.  This means that other illegal provisions would be open to judicial redrafting.  In my opinion, the availability of “notional severance” as a remedy creates greater uncertainty in the law.  It is clear that both severance and notional severance alter the parties’ agreement in some way.  However, under the traditional severance approach, courts continue to work with the words of the parties themselves, removing only those portions of the contract that render it illegal.  In stark contrast, under notional severance, courts will be permitted to literally add new words to the parties’ agreement.  By doing so, courts will be substituting their intentions for those of the parties.  Notional severance would extend the judicial role in what I consider to be an unfortunate way.  One instance, in particular, comes to mind.

60                               In Canadian American Financial Corp. (Canada) Ltd. v. King, supra, Hinkson, Lambert and Southin JJ.A. wrote three sets of concurring reasons refusing to grant an interlocutory injunction to enforce a covenant in restraint of trade. The provision of the contract in question was a non-competition clause that restrained competition in Canada and Bermuda.  All three judges held that the covenant was overly broad.  They refused to substitute a covenant with reasonable terms for the unreasonable provision.  Lambert J.A. concluded that the authorities were clear that courts may not make contracts for parties that they have not made themselves.  In his opinion, substituting the words “British Columbia and Alberta” for “Canada and Bermuda” would have rendered the term enforceable (p. 307), but that was something for the parties, not the court, to do.  Under the new approach, would judges be permitted to make such substitutions in order to render the contract enforceable? 

61                               It is important to note that Lambert J.A. found, at p. 307, that the blue-pencil rule would have made no difference to his decision as he would not have been any more willing to sever a portion of the clause had it named all the provinces and territories of Canada separately, and so been amenable to the application of the blue-pencil rule.  (I am not convinced under traditional severance that a court would not be permitted to strike out a number of offending provisions, or a number of provinces within a list, in order to render the contract compliant with the law.  As the majority’s decision preserves traditional severance as one of the remedies within the trial judge’s discretion, this issue will likely be considered in a more appropriate case in the future.)

62                               Furthermore, Lambert J.A. held that a clause, where one alternative encompasses another, would be, in his opinion, void for uncertainty and should not be made valid by severance. For example, a contract containing a covenant not to compete:  (1) in Ottawa, and (2) in the rest of Ontario, might be valid as to Ottawa, but overbroad, and therefore wholly invalid, as to Ontario.  Including the two promises within the same contract renders the whole clause invalid.  In Lambert J.A.’s opinion, the only point on which the covenant is clear is that the covenant is to cover at least Ottawa.  But the parties could have simply stated that.  Again, under the approach adopted by my colleague, could a court rewrite the second clause to read, for example, “in the rest of northeastern Ontario”, so that it no longer constitutes an illegal restraint of trade?

63                               Third, in my view, the approach taken by the majority in this case is inconsistent with that taken in Garland v. Consumers’ Gas Co., [1998] 3 S.C.R. 112.  In Garland, this Court interpreted the definition of interest broadly in order to prevent creditors from avoiding the statute by manipulating the form in which payments are to be made.  At paras. 27‑28, Major J. wrote:

It is apparent from this definition that for the purposes of s. 347 “interest” is an extremely comprehensive term, encompassing many types of fixed payments which would not be considered interest proper at common law or under general accounting principles. . . .

. . .The broad language of s. 347 was presumably intended . . . to prevent creditors from avoiding the statute simply by manipulating the form of payment exacted from their debtors . . . .  It is the substance, and not merely the form, of a charge or expense which determines whether it is governed by s. 347.

64                               Under my colleague’s approach, a creditor would be permitted to escape the consequences of its avoidance measures by simply reducing the rate applied to the maximum permitted under the Criminal Code . In my opinion, this approach is inconsistent with the general objectives expressed in the Code and incompatible with the notion of deterrence.  It is a criminal offence for any lender, even a commercial creditor, to enter into a loan agreement providing for an effective interest rate greater than 60 percent per annum. And, as Rosenberg J.A. remarked, at para. 31, of the majority reasons of the Court of Appeal ((2002), 60 O.R. (3d) 97) to permit a notional severance of the kind ordered by the application judge is to effect a substantial innovation in the common law doctrine of severance to the benefit of those who prima facie stand in violation of the criminal law.

65                               It is also important to note that this Court has developed a method for determining whether, as a matter of principle, a departure from well-established precedent is warranted. I canvassed the relevant factors in Friedmann Equity Developments Inc. v. Final Note Ltd., [2000] 1 S.C.R. 842, 2000 SCC 34, at para. 43, as follows:

In the recent case of Robinson [R. v. Robinson, [1996] 1 S.C.R. 683], Lamer C.J., for a majority of the Court, relied on five factors to justify the reversal of an earlier decision of the Court in MacAskill v. The King, [1931] S.C.R. 330.  These factors were the existence of previous dissenting opinions in this Court, a trend in the provincial appellate courts to depart from the principles adopted in the original decision, criticism of the case or the adoption of a contrary rule in other jurisdictions, doctrinal criticism of the case and its foundations, and inconsistency of the case with other decisions. While they are not prerequisites for a change in the common law, these factors help to identify compelling reasons for reform.  On the other hand, courts will not intervene where the proposed change will have complex and far-reaching effects, setting the law on an unknown course whose ramifications cannot be accurately measured:  see Bow Valley [Husky (Bermuda) Ltd. v. Saint John Shipbuilding Ltd., [1997] 3 S.C.R. 1210], at para. 93.

In my opinion, there do not appear to be compelling reasons to depart from the lower courts’ and Privy Council’s well-established precedent.

66                               My colleague has provided a comprehensive summary of the facts in issue.  In my opinion, the finding by the trial judge that there was no criminal intent should not cloud our analysis.  Knowledge is not a requirement under the provision.  Indeed, as noted by Rosenberg J.A., at para. 15 of the majority reasons of the Court of Appeal, the typical case is one where the parties do not intend to violate the criminal law and occupy relatively equal bargaining positions. 

67                               Also, while legal advice will almost certainly be helpful, parties cannot rely on such advice to excuse themselves from non-compliance with the law.  Here, advice was given on the specific interest clause only, that the interest was 48 percent, and that advice was wrong.  Neither party was advised by its solicitor of the scope of s. 347  of the Criminal Code  (see appellant’s factum, at para. 5; respondent’s factum, at para. 1).  Contracting parties should consider all of the provisions in the agreement when determining the total interest charged.

68                               I would like to conclude by drawing attention to clause 9.1 of the accounts receivable factoring agreement which reflects the parties’ own intentions as to remedy (see tab 4 of the respondent’s record).  Clause 9.1 reads as follows:

9.1  Severability.  If any provision of this Agreement is deemed by a court having jurisdiction to be wholly or partially invalid, this Agreement shall be interpreted as if such provision had not been part of this Agreement, and so that the validity of such provision shall not affect the validity of the remainder of this Agreement which shall be construed as if this Agreement had been executed without such provision.

69                               At the hearing of the appeal in this Court, counsel for the appellant submitted that “this severance provision really doesn’t apply to the interest rate provision that is in effect in this case which is found in a separate document which was in the commitment letter”.  Counsel conceded however that clause 9.1 “reflects an intention that if, to the extent that a provision is tainted by illegality, it will not render the contract invalid, and this clause applies to the extent that notional severance is not effected” (emphasis added).  The underlined caveat is, of course, incompatible with the plain meaning of clause 9.1 and unsupported by the record. 

70                               In any event, the commitment letter referred to by counsel is dated March 6, 2000, and the factoring agreement that contains clause 9.1 indicates that it was executed as of March 30, 2000. And, significantly, the commitment letter and the factoring agreement provide for the same royalty payment of $160,000, the same monitoring fee of $750 per month, and the same standby fee of 1 percent per month.

71                               A letter sent from the appellant’s solicitor to the respondent is instructive in this regard. As the trial judge explains, at paras 4‑5 of his reasons ((2001), 54 O.R. (3d) 144):

Although a factoring agreement, promissory note and general security agreement were executed on — or as of — March 30, 2000, the parties had already agreed to depart from the provisions of the first of these documents. On March 28, 2000, the respondent’s solicitor wrote to the solicitor acting for the applicant and the Dragosits:

Further to our conversation of March 27, 2000, I confirm on behalf of my client that as an economic accommodation to your client, the administration of the factoring facility will not strictly follow the provisions of the factoring agreement, unless my client elects to exercise its rights under that agreement. It is anticipated that your client will draw the full amount of the facility and pay interest, fees and royalties as they are set out in the commitment letter. Please note that the interest rate [is] 48 (per cent) per annum.

In consequence, the concept of a factoring of receivables was put aside and replaced by a revolving credit facility secured on the current assets of the applicant including its accounts receivable. Each of the Dragosits signed a personal guarantee for the entire $500,000.  [Emphasis added.]

72                               In these circumstances, I am not persuaded that clause 9.1 of the factoring agreement was in any way overtaken by the commitment letter dated earlier.  While certain aspects of the financial arrangement were governed by the commitment letter alone, it is apparent from the letter sent by appellant’s solicitor that the provisions of the factoring agreement were still very much alive.

73                               A plain reading of the clause indicates that, should a term of the agreement between the parties be “deemed . . . wholly or partially invalid”, the parties intended their agreement to be “construed as if this Agreement had been executed without [the offending] provision”.  I cannot see how the wording of the clause can be said to anticipate — still less, to provide for — notional severance.

74                               Bearing in mind the plain meaning of clause 9.1, counsel’s concession as to the intention of the parties and the letter sent by the appellant’s solicitor to the respondent on March 28, 2000 —  three weeks after the commitment letter, on which the appellant relies, was signed, I feel bound to conclude that the blue-pencil approach adopted by the Court of Appeal is entirely consistent with the parties’ own intentions and that notional severance is not.

75                               In light of this consideration, I respectfully disagree with my colleague that reducing, or effectively rewriting, the rate of interest to the legally prescribed rate of 60 percent would show more respect for the sanctity of the bargain between the parties.  It appears from clause 9.1 that the parties turned their minds to this very issue and agreed that, in the event of a court declaring a provision in the contract invalid, the agreement shall be interpreted, not redrafted, as if that provision was not part of the contract. 

76                               For these reasons, I would dismiss the appeal with costs.

The reasons of Deschamps and Fish JJ. were delivered by

77                               Fish J. (dissenting) — With respect for the contrary view of Justice Arbour, I would, like Justice Bastarache and for the reasons that follow, dismiss the appeal with costs.

I

78                               We are concerned in this case with the civil consequences of a serious violation of the criminal law of Canada.

79                               Section 347  of the Criminal Code , R.S.C. 1985, c. C‑46 , provides that it is a criminal offence to charge or agree to receive “interest at a criminal rate”, which is defined as interest that exceeds 60 percent per annum.  I consider the violation in this case to be serious because the appellant charged the respondent an effective interest rate of more than 90 percent.

80                               I recognize that the law of equity is not a closed-end list of off-the-shelf remedies.  Like the majority in the Court of Appeal, however, I do not believe this case required a new and novel prescription to ensure a fair and reasonable solution.  The well-established “blue-pencil” remedy described by my colleagues and applied in the Court of Appeal respects the trial judge’s findings of fact, as it must, and achieves an equitable result consistent with established principle.

81                               Finally, I agree with the majority in the Court of Appeal that the trial judge erred in straining the recognized rules of equity by resorting to what he described as “notional severance”.  In different circumstances, his fresh and creative approach might well prove appropriate. That would be so where notional severance is necessary to resolve a private dispute fairly and in a manner that is not incompatible with the social and legal objectives of the criminal law — which, in my respectful view, is not the case here.

II

82                               The appellant made a commercial loan to the respondent at a “criminal rate” within the meaning of s. 347  of the Criminal Code .  Pursuant to that provision, as I have mentioned, anything over 60 percent per annum is, as a matter of law, a “criminal rate”. 

83                               The loan in this case was made at an effective annual rate of 90.9 percent, or 150 percent of the legal maximum.

84                               There were two components to the interest charged. 

85                               The first provided for interest at four percent per month, calculated daily, payable monthly in arrears. Appellant’s solicitor, without intending to mislead, erroneously advised respondent’s solicitor “that the interest rate [was] 48 (per cent) per annum”. It was in fact 60.1 percent, a criminal rate in itself.

86                               The second component consisted in various charges that, taken together, increased the effective annual interest rate from 60.1 percent to 90.9 percent.

87                               These facts are not in dispute.

88                               Moreover, the appellant, though it did so at the outset, no longer contests: (1) that the rate stipulated as interest was a “criminal rate” within the meaning of s. 347  of the Criminal Code ; and (2) that the additional charges I have mentioned fall within the definition of “interest” in s. 347.  And it is common ground that s. 347 applies to the kind of commercial loan that concerns us here.

89                               Two defining features of the impugned loan agreement must therefore be borne in mind from the outset.  First, that we are not dealing in this case with a contract that violates civil, commercial or regulatory legislation, but rather with an agreement that violates the Criminal Code  of Canada ; and second, that the effective interest rate under that contract amounts to more than one-and-a-half times the threshold criminal rate, and does not exceed it barely, slightly, or by only 0.1 percent.

90                               The trial judge found that the appellant did not intend to breach s. 347  of the Criminal   Code .  This was therefore, in his view, a case for severance of the illegal provisions in the contract.

91                               It is important to remember that only one provision of the contract was illegal in itself:  the provision which charges interest at the criminally prohibited rate of 60.1 percent.  The other charges, included in the definition of interest set out in s. 347, were separately set out in the agreement and not stipulated as interest.

92                               Had the trial judge severed the interest clause and left the other charges intact, his decision would have fallen within the bounds of his discretion.

93                               He declined, however, to do so. 

94                               Instead, he applied a new and novel remedy that he described as “notional severance”, rewriting the interest clause so as to yield, when added to the other charges, a cumulative, effective interest rate of 60  percent — the very threshold of the criminal rate prohibited by s. 347  of the Code .

95                               In the Court of Appeal for Ontario, the majority found that the trial judge had erred in “straining” the law in this way.  The appellant had charged the respondent a criminal rate of interest far exceeding the limit established by Parliament, and the Court of Appeal found it inappropriate, as a matter of legal principle and judicial policy, to bend or extend the equitable remedy of severance in order to secure for the appellant the maximum rate of interest it could have recovered had it not violated the law. 

96                               Bearing in mind the trial judge’s findings of fact and applying the recognized blue-pencil approach, the Court of Appeal instead severed the criminal rate agreed to by the parties as interest and permitted the appellant to recover the separately stipulated charges, which left in place an effective interest rate of just over 30 percent.

97                               The decisive question in this Court is whether the Court of Appeal erred in law in concluding as it did.

98                               I would answer that question in the negative. 

III

99                               Without so deciding, I am prepared for present purposes to agree with Arbour J., at para. 5, that “notional severance is available as a matter of law as a remedy in cases arising under s. 347 ” of the Criminal Code .

100                           Even on that assumption, however, I would permit notional severance only where: (1) public policy does not require that the entire agreement be declared unenforceable; (2) severance is found to be warranted; and (3) severance simpliciter — or “blue-pencil severance” — is impracticable or would occasion an unjust result.

101                           Here, on the facts as found by the trial judge, the first two criteria are satisfied but the third is not: blue-pencil severance is both possible and fair.  And, unlike notional severance as applied by the trial judge, blue-pencil severance does not do violence to the policy purposes of s. 347  of the Criminal Code  or require a judicial rewriting of the interest clause agreed to, as such, by the parties.

102                           Section 347(1)  of the Code  provides that everyone who

(a)  enters into an agreement or arrangement to receive interest at a criminal rate, or

(b)   receives a payment or partial payment of interest at a criminal rate,

is guilty of

(c)  an indictable offence and is liable to imprisonment for a term not exceeding five years, or

(d)  an offence punishable on summary conviction and is liable to a fine not exceeding twenty-five thousand dollars or to imprisonment for a term not exceeding six months or to both.

103                           And s. 347(2) provides:

(2) In this section,

. . .

“criminal rate” means an effective annual rate of interest calculated in accordance with generally accepted actuarial practices and principles that exceeds sixty per cent on the credit advanced under an agreement or arrangement;

104                           As consideration for its loan to the respondent, in addition to the interest stipulated as interest, the appellant, as already mentioned, charged various other amounts that constituted interest called by another name.  As noted by Arbour J., it is undisputed that those charges fall squarely with the definition of “interest” set out in s. 347(2)  of the Criminal Code .  Together, they added 30.8 percent to the 60.1 percent expressly contemplated by the agreement.

105                           The trial judge found as a fact that the appellant, in charging the respondent an effective interest rate of 90.9 percent per annum, did not intend to breach the provisions of s. 347.  His conclusion in this regard should not be misunderstood: an intent to breach the section is not an essential element of the offence.

106                           There is no doubt, moreover, that the appellant was perfectly aware of the stipulated rate of interest (though perhaps not of its actuarial quantification) and was aware of all the other charges that, as a matter of law, constitute interest as well.

107                           As I said earlier, I am prepared to assume for present purposes that notional severance is a remedy known to law, but would permit its application in cases involving a breach of the criminal law only where severance is nonetheless warranted as a matter of equity and severance simpliciter is impracticable or would create an unjust result.

108                           The question, then, is whether this is such a case. 

109                           With the greatest of respect, I am persuaded that it is not, essentially for the reasons set out by Rosenberg J.A., speaking for the majority in the Ontario Court of Appeal ((2002), 60 O.R. (3d) 97, at paras. 31‑35):

It is a criminal offence for any lender, even a commercial creditor, to enter into a loan agreement providing for an effective interest rate greater than 60 per cent per annum.  And to permit a notional severance of the kind ordered by the application judge is to effect a substantial innovation in the common law doctrine of severance.  While the common law develops over time, it would be inconsistent with the aims of deterrence for the courts to make such a major innovation at the behest of those who prima facie stand in violation of the criminal law.  As Huddart J. [now J.A.] said in Pacific National Developments [Ltd. v. Standard Trust Co. (1991), 53 B.C.L.R. (2d) 158 (S.C.),] at p. 163 . . . :  “It is not for the courts either to strain to find ways in which to avoid the application of the Criminal Code , or to assist those who strain to do so, although it is for the courts to interpret strictly penal provisions of statutes”.

. . .

I agree with the British Columbia Court of Appeal in Terracan, supra, that a judge has discretion to apply the doctrine of severance to an agreement that offends the criminal interest rate provisions of the Code .  This discretion gives rise to a spectrum of available remedies.  Where the loan transaction resembles a traditional loan sharking arrangement, the court may refuse to apply the doctrine of severance and hold the entire loan agreement unenforceable, including the obligation to repay the principal.  While this remedy leaves the borrower with a windfall, this result may be justified in some cases by the need to denounce such usurious practices.  See C.A.P.S. International Inc. v. Kotello, [2002] M.J. No. 205 [(QL)] (Q.B.).  At the other end of the spectrum, in the case of a good faith commercial transaction where the equities favour the lender and severance does not undermine the policy of the legislation, the court may sever only those provisions of the loan agreement that put the effective interest rate over 60 per cent, leaving intact the borrower’s obligation to repay the principal and pay some interest.  See e.g., Milani, supra.  Closer to the centre of the spectrum lies a case like Terracan, where the court severed all the interest provisions but upheld the debtor’s obligation to repay the principal.

Where does the present case fit on the spectrum?  The application judge concluded, and I have accepted, that the equities favour the lender.  His reasons make it clear that if he had concluded that notional severance was not available, he would have severed the provision calling for payment of an interest rate of 4 per cent per month calculated daily, payable monthly in arrears.  I agree that this is the appropriate remedy. 

In this case there are at least two distinct covenants that can be traced back to the negotiations leading to the agreement.  As indicated, the respondent initially wanted an equity interest in the appellant’s business.  This is the source of the royalty payments.  The other covenant was the more conventional agreement with respect to interest.  In my view, it is possible to sever one or both in a manner consistent with the established principles respecting severance.  The 4 per cent per month interest provision must be severed because it violates s. 347  of the Criminal Code .  I see no justification for trying to save that provision simply because it only barely exceeds 60 per cent.  The respondent is an experienced commercial lender.  It misled the appellant, albeit unintentionally, when its solicitor referred to the interest rate as merely 48 per cent.  The appellant has not shown that the other fees and charges such as the commitment fee and monitoring fee are clearly part of the interest rate payment and accordingly they should not be severed.

110                           I agree with Rosenberg J.A.’s analysis and would add these additional considerations. 

111                           The first relates to the appellant’s purported good faith.  As I understand it, this refers to:

(1)  the trial judge’s finding that the appellant did not intend to breach s. 347, in the sense that the appellant appears not to have realized that its otherwise well-understood charges violated the Criminal Code ; and

(2) the fact that the separate and distinct charges which added just over 30 percent per annum to the stipulated rate of interest were not known by the appellant to constitute interest within the meaning of s. 347(2)  of the Code .

112                           I accept that neither the appellant nor the respondent thought of the added charges as interest.

113                           On this assumption, it appears to me neither “artificial” nor “arbitrary” to sever the criminal rate of interest agreed to as interest and to leave intact the distinct and separate charges not agreed to as interest, and not considered to be interest by either party.

114                           This solution responds adequately to the policy concerns expressed by the majority in the Court of Appeal.  Unlike notional severance, it requires no judicial rewriting of the interest clause agreed to as such by the parties.  And it still leaves the appellant with a return of slightly more than 30 percent per annum.

IV

115                           Arbour J. concludes that the four factors set out by Blair J.A. in William E. Thomson Associates Inc. v. Carpenter (1989), 61 D.L.R. (4th) 1 (Ont. C.A.), favour severance in this case.  I agree.  That is why I would affirm the judgment of the Court of Appeal.

116                           But the more difficult question is whether these four factors favour notional severance as opposed to the established blue-pencil form of severance contemplated by Blair J.A. in Thomson.  In my respectful view, they do not.

117                           The first consideration or factor is whether the purpose and policy of s. 347  of the Criminal Code  would be better served by severance simpliciter or by notional severance.

118                           In support of her conclusion, on this branch of the matter, Arbour J. cites at para. 43, an article in which Professor Ziegel explains his objections to s. 347 (J. S. Ziegel, “Bill C‑44: Repeal of the Small Loans Act and Enactment of a New Usury Law” (1981), 59 Can. Bar Rev. 188). On the issue that concerns us here, however, Professor Ziegel elsewhere states: 

So far as the interest component is concerned, obviously the lender cannot recover that part of it which exceeds 60%.  But can he recover the first 60%?  In my view, the answer should be no, and this on two grounds.  First, it is unlikely that the contract itself will distinguish between the two interest components.  There is therefore no basis on which to apply the doctrine of severance.  The second reason is that severance would undermine the policy of s. 305.1 [now s. 347] (even if one believes, as the writer does, that it is a bad policy) and encourage lenders to run calculated risks, secure in the knowledge that they would only lose that part of the interest which exceeds 60%.  [Emphasis added.]

(J. S. Ziegel, “The Usury Provisions in the Criminal Code : The Chickens Come Home to Roost” (1986), 11 Can. Bus. L.J. 233, at p. 242)

119                           Professor Ziegel’s first reason is of limited application here, since the parties did distinguish between the two components of the illegal interest rate.  Moreover, there is a sound basis for distinguishing between them:  the first, agreed to by the parties as interest, itself exceeds the limit, albeit narrowly, while the second was not considered by the parties to be part of the interest rate, though s. 347 gives it that effect.

120                           Professor Ziegel’s second reason, which I share, clearly militates in favour of blue-pencil severance, the solution adopted by the majority in the Court of Appeal, and against “notional severance”, which the trial judge applied in awarding the appellant interest at the threshold of the criminal rate.

121                           I agree with Arbour J. that the second consideration — whether the parties entered into the agreement for an illegal purpose or with an evil intention — in this case favours a flexible remedy.  That objective, however,  is satisfied by severing the interest clause and leaving the other charges intact, since this requires the respondent to pay, and allows the appellant to receive, an effective interest rate of more than 30 percent on the commercial loan that was the object of their agreement. 

122                           This remedy certainly differs from — but is no less “flexible” — than the remedy conceived by the trial judge. 

123                           Had there been only an effective interest rate of 60.1 percent — and not a cumulative effective rate of 90.9 percent — it might have been desirable, I concede, not to sever the interest clause entirely.  But that was not the case here.

124                           The third consideration relates to the relative bargaining positions of the parties and their conduct in reaching the agreement.

125                           In this regard, it is true that the respondent had the advice of a solicitor and does not appear to have been a naive or inexperienced victim of misunderstanding.  But a commercial borrower who agrees, for lack of an alternative source of financing, to pay an effective interest rate of 90.9 percent per annum — many times the normal commercial rate — and to provide personal guarantees for the loan, can hardly be said to have bargaining power equal to that of the lender.

126                           I turn now to the fourth consideration, relating to the potential for unjust enrichment or, in the words of Arbour J., an “unjustified windfall” (para. 42).

127                           Unlike Arbour J., I do not consider that the respondent can be said to be unjustly enriched if it is relieved by a court of law of its contractual undertaking to pay a criminal rate of interest where, as here, it is required at the same time to pay an effective annual rate of more than 30 percent and to repay the principal as well — in full.

128                           In short, it appears to me that all four considerations identified by Blair J.A. in Thomson, supra, examined individually and weighed together, militate here  in favour of the remedy applied by the Court of Appeal and against the remedy adopted by the trial judge.

129                           Accordingly, I agree with the Court of Appeal that the trial judge erred in substituting for the criminal rate of interest charged by the appellant to the respondent the highest rate the appellant could otherwise have charged. 

130                           And, like the Court of Appeal, I see no basis in the established rules of equity that would permit the trial judge to rewrite the illegal interest clause agreed to by the parties in this way.  It should simply have been struck out on the accepted blue-pencil approach.

131                           The effect of the trial judge’s decision was to stretch the principles of equity in an inappropriate way and to send the wrong message to those who lend money at a criminally prohibited rate to “willing” borrowers who cannot otherwise obtain a loan.

132                           They should not be encouraged to believe that if their illegal arrangement is subjected to judicial scrutiny, they will nonetheless recover the highest rate they could legally have charged — and thus suffer no pecuniary disadvantage for having violated s. 347  of the Criminal Code .

V

133                           As mentioned at the outset, I have assumed for the purposes of this appeal  the correctness of Arbour J.’s premise that “notional severance is available as a matter of law as a remedy in cases arising under s. 347 ” of the Criminal Code .

134                            On this assumption, I would apply notional severance to cases involving a breach of the Criminal Code  only where severance does no violence to public policy interests, is found to be warranted as a matter of equity, and severance simpliciter is impracticable or, though possible, would create an unjust result.

135                           For all of these reasons and, to the extent that they are not incompatible, the reasons of Bastarache J., I would dismiss the appeal with costs.

Appeal allowed with costs, Bastarache, Deschamps and Fish JJ. dissenting.

Solicitors for the appellant:  Fraser Milner Casgrain, Toronto.

Solicitors for the respondent:  Ackerman Law Office, Toronto.