"Pro-Competitive Executive Compensation" as a Condition for Approval of Mergers that Simultaneously Exploit Consumers and Enhance Efficiency

(19 Canadian Competition Record 18 (Spring 1998)

By: Ian Ayres, Yale Law School/University of Illinois College of Law (Visiting) and

Stephen F. Ross, University of Illinois College of Law

The pending mergers between the Royal Bank of Canada and the Bank of Montreal, and Toronto-Dominion Bank and the Canadian Imperial Bank of Commerce, raise complex issues in the application of sound competition policy. Although economists have identified the potential trade-offs between competition and efficiency,⁽¹⁾ the larger size of the United States markets has traditionally allowed American antitrust policymakers the luxury of condemning mergers that threatened to facilitate oligopoly pricing, secure in the view that most economies of scale could be achieved by internal growth or through acquisitions that did not significantly raise concentration.⁽²⁾ Canadian competition policy, in contrast, has long been sensitive to the potential for mergers to increase market power due to increased concentration simultaneously with achieving significant efficiencies in the deployment of Canadian economic resources, as reflected in section 96 of the Competition Act, which directs the Competition Tribunal to forego a merger-blocking decree where efficiencies will offset harms to competition.

In Philadelphia National Bank v. United States, ⁽³⁾ the United States Supreme Court held that U.S. antitrust law prohibited the merger of the first and third largest commercial banks in Philadelphia. The critical aspect of that case for American competition policy was that the perceived exploitation of ordinary consumers and small businesses in the Philadelphia metropolitan area could not be justified on the ground that the merging firm could achieve a scale sufficient to compete with major New York "money center" banks for the patronage of Philadelphia's largest corporations. This same policy issue directly confronts Canadian officials evaluating the pending bank mergers.

With what we hope is the appropriate humility of outsiders offering advice to our northern neighbors, we propose a novel approach to mergers that simultaneously create efficiencies and market power. Specifically, we propose that approval of the mergers be conditioned on undertakings by the merged firms that a significant portion of the compensation for senior executives will be tied to the relative position of their own companies vis-a-vis their principal rival. These "pro-competitive compensation packages" could take the form of significant bonuses based on improvements in relative market shares, or of financial instruments that give, for example, senior management at Royal Bank a strong financial incentive not simply to maximize the value of Royal Bank stock, but to maximize its stock vis-a-vis in relation to the stock value of Toronto Dominion, and vice versa. We believe this approach will allow policymakers to permit the firms to take advantage of efficiencies of size and scope with some confidence that the efficiencies will rebound to the benefit of Canadian consumers.

Supporters of the proposed bank mergers have asserted that the acquisitions will result in substantial cost savings, in large part from the ability of the merged entities to serve banking customers with substantially reduced capacity in terms of physical locations, due to advances in technology. Supporters and opponents disagree, however, about whether the acquisitions are necessary for each firm to respond to changing technology, and whether new entry will be sufficient to check the increased opportunity for unilateral exercise of market power or coordinated interaction among the remaining Canadian banks.

We are not aware of any serious claim by supporters of the two pending mergers that the government should approve acquisitions that actually result in the economic exploitation of ordinary Canadian consumers so that the resulting monopoly profits may be "wisely invested" by the two remaining superbanks in order to compete more effectively in the global market. Nor have there been any serious claims made, like the ones made by the defendants in Philadelphia National Bank, that the harm to ordinary Canadians will be substantial but nonetheless it is counterbalanced by benefits to major Canadian corporations who must compete globally, because these corporations, with the ready option of turning to foreign banks, will clearly see the benefits of any efficiencies passed on to them. Thus, conventional approaches to merger enforcement require a judgement as to whether to take advantage of the efficiencies or to prevent consumer exploitation.

Of course, the optimal result would be one that would allow banks to take full advantage of available efficiencies while maintaining or even increasing vigorous domestic rivalry for the benefit of ordinary Canadians. Unless, however, the Competition Bureau concludes that new entry will really be sufficient to prevent the unilateral exercise of unilateral market power or interdependent behaviour⁽⁴⁾ among remaining firms, merger approval provides no assurance that a reduction in the number of major national banks from five to three will not significantly impair the price or quality of banking services for ordinary Canadians.⁽⁵⁾ Although we do not presume to replicate the complex merger analysis that the Bureau will undertake, we note that the Bureau's general approach to bank mergers suggests the likelihood that there will be some localized markets where a finding of lessening of competition will be made.⁽⁶⁾ Unless the Bureau concludes that the proposed mergers are unlikely to result in significant efficiency gains, traditional merger analysis requires the Bureau to engage in the difficult and high-stakes determination of whether the efficiencies outweigh the harm to Canadians. We submit that our proposal presents the Government with an attractive alternative likely to achieve a superior result.

Still-accepted economic wisdom is that executives at Royal Bank, charged with maximizing shareholder value and typically provided with stock and/or stock options to give them a personal incentive to achieve that goal, will seek, where possible, to exercise unilateral power over some consumers, and to engage in interdependent bahaviour with Toronto Dominion vis-a-vis other consumers. On the other hand, were Royal Bank executives' compensation linked to the relative performance of their bank vis-a-vis their chief rival, there would be a strong incentive not to engage in coordinated interaction, but to aggressively compete in current markets as well as to enter new markets to prevent any exercise of unilateral power by Toronto Dominion.

Our core idea is that as a condition of merging, regulators should require compensation packages that induce managers to increase their profitability by out-competing their rivals, rather than their profitability by tacitly or explicitly colluding with their rivals. Traditional stock incentive plans give managers an incentive to raise a firm's profits, but in an oligopolistic industry this increased profitability may come merely from supra-competitive pricing across the industry as a whole. While shareholders are indifferent about how their managers increase profits, competition regulators should not be.

Although the potentially divergent interests of managers and shareholders is normally thought of as an "agency problem," we envision exploiting the agency relationship to enlist the managers to eschew the kinds of explicit or tacit collusion that would benefit their shareholders but hurt society generally. There are myriad of ways to alter managers' incentives in a socially beneficial manner. One, which we call the "market share" approach, would be to define economically relevant markets in which the Bureau is concerned that post-merger competition will be insufficient, and require a significant portion of compensation for dominant firm managers to come in the form of contractually-agreed upon bonuses for increases in market share over their principal rival. For example, suppose that at the beginning of the Royal Bank ("RB") fiscal year, RB's average market share in individual checking accounts in geographic markets with less than one million people⁽⁷⁾ is 45%. Under this approach, for example, RB would be required to pay a manager whose target compensation package was $1 million with $800,000 in salary and a $50,000 bonus for every percentage point its market share exceeded 41%. If the RB's market share remained constant at 45%, the manager would earn a $200,000 bonus.⁽⁸⁾ But tying the manager's bonus to the bank's relative market position would give the manager a strong incentive to cut price (and/or increase the quality of its product) to lure customers away from its rivals.

A similar price cutting incentive could be produced by a variety of other bonus schemes that turn on the quantity (say, of checking accounts) produced.⁽⁹⁾ For example, in the foregoing example, if Royal's and Toronto Dominion's ("TD") market share were initially 45% and 35%, respectively, it might be useful to require that RB's bonus be paid $50,000 for every percentage point over 6% that RB's share of the market exceeded TD's at year end. If the market shares remained unchanged, RB's manager again would earn a $200,000 bonus. But since this bonus tied to RB's market share relative to the share of its largest rival, RB's manager will have an incentive to focus on strategies that lure away TD's customers instead of possibly gaining market share by predating on fringe firms.

"Market share" bonuses might lead managers to price too aggressively -- cutting the price of their financial products below cost merely to increase the size of their bonus. Accordingly, it would be advisable to condition the market share bonus on some minimal level of profitability and/or to allow the firm to combine the market share bonus with traditional stock option plans which cause bank managers to trade-off the effects of increased output and decreased firm value. The critical point is that when regulators are worried that a merger will lead to a (possibly artificial) reduction in output, they can require that managers be given incentives that work in just the opposite direction.

An alternative approach, which we call the "financial instrument" approach, is a bit more complicated to explain but potentially easier to administer. This approach would require merged firms to give their managers a balanced portfolio of calls on their own stock and puts on their rival's stock to help ensure that managers have an incentive to compete and not collude.

To be specific, we imagine requiring the merged banks to give their managers at the beginning of each fiscal year a bundle of puts in their rival's⁽¹⁰⁾ stock. The puts would have a year duration (and could only be exercised by the managers at the end of the year). Each year the firm would need to give their managers another bundle of puts. The puts would be issued "in the money" -- meaning that the exercise price of the manager's option to sell would be above the current stock price of the rival bank. To be effective, the compensation plan should conform to two general principles.

First, the merged banks should be required to give their managers enough "in the money puts" of their rival's stock so that if stock prices didn't change the managers would receive 10% of their total compensation from cashing in the puts. For example, suppose Royal Bank want to provide a senior executive with $1,000,000 in compensation for a particular year, and at the beginning of the year both its stock price and Toronto Dominions' are trading at $100 per share. Royal could meet the 10% requirement by paying the manager $900,000 in cash together with 5000 puts in its rivals with an exercise price of $120. If at year's end, the rival's stock price remains at $100, the manager would be able to sell the puts for $100,000 profit.⁽¹¹⁾

Second, the manager's total compensation must be at least as sensitive to decreases in the total market value of its rival's equity as it is to increases in the total market value of its own equity. Currently, many managers are compensated with call options in their own corporation's stock, and we believe this incentive will be intensified if a corporation is required to give its managers puts in the stock of its rival. But if stock or option incentives in its own stock dwarfed the put options in the rival's, managers might be inclined to ignore the put options and simply make money by engaging in the interdependent behaviour that our proposal is designed to deter.

To provide a stylized example of how these two principles work to give managers an incentive to compete instead of an incentive to collude, suppose both RB and TD each have a total stock market value of $100 million with one million shares outstanding, so that stock price is $100/share. Under our proposal, a manager might be paid $800,000 in cash, plus 5000 calls in RB stock with an exercise price of $80, and 5000 puts in TD stock with an exercise price of $120.

This combination of puts and calls significantly weakens the manager's incentive to collude. If cartelization would raise the market value of both RB and TD stock $120, the extra compensation the manager would gain on its calls ($100,000) would be exactly offset by the amount that the manager would lose on its puts ($100,000). The manager's total compensation would remain at $1,000,000, so the effort of colluding would not increase the manager's personal gain. The only way RB executives can increase their compensation above $1,000,000 is to cause RB stock value to exceed TD's value. As shown on the following page, if an RB manager can cut fixed costs and thereby increase RB's profitability and share price (say to $110), her compensation would increase (to $1,050,000). But the manager profits even more if it can cause RB stock price to increase at the expense of TD. If RB's stock rises to $110 while TD stock falls to $90, the manager's total compensation increases even more to $1,100,000.⁽¹²⁾

Analyzing Different Scenarios Under Put/Call Pro-competitive Compensation Plan

Assumptions:

$800,000 in cash

5000 "in the money" calls on Royal at $80

5000 "in the money" puts on Toronto Dominion at $120

Manager's Total Compensation as a function of different year end prices:

Both the "market share" and the "financial instrument" approaches give bank managers a heightened incentive to chisel on tacit agreements to price supracompetitively. By cutting price (say on checking account or loan fees), the "incentified" manager could shift sales from the rival to itself and hence increase its bonus. The essence of these "pro-competitive" compensation plans is that on the margin, they make it more profitable for managers to compete than to collude. These plans increase a manager's compensation when her employer does relatively better than its rivals, which is precisely what competition is about. The Competition Act, like U.S. antitrust law, embraces the view that "immunity from competition is a narcotic, and rivalry is a stimulant to industrial progress; that the spur of constant stress is necessary to counteract an inevitable disposition to let well enough alone."⁽¹³⁾ Whatever the merits, in ordinary social discourse, of the "Shaudenfreude" principle of taking joy from the suffering of others, the Competition Act is designed to encourage a process where competitors dislike each other and "almost always want to hurt each other's business."⁽¹⁴⁾

To the extent that, with five national banks dominating the market, consumers are already victimized by some degree of interdependent bahaviour, we concede that our proposal may well result in even greater benefits to Canadian consumers than currently exist. Although we are inclined to think that this aspect of our proposal is a strength, rather than a weakness, we are cognizant that Canadian law only permits antitrust intervention to restore the pre-merger level of competition.⁽¹⁵⁾ However, the Supreme Court of Canada has made it clear that if "the choice is between a remedy that goes farther than is strictly necessary to restore competition to an acceptable level and a remedy that does not go far enough even to reach the acceptable level, then surely the former option must be preferred."⁽¹⁶⁾ If economic science were precise enough to assess the cost and demand curves in various banking markets in order to target, with any confidence, our pro-competitive compensation proposal to precisely the level of competition that existed premerger, Canada wouldn't need an anti-merger policy at all; with this sort of omniscience, the government could simply regulate the post-merger price directly! Without this sort of precise data, we believe our proposal to be the one most likely to be "effective," as the courts require.

Moreover, the concern about overly aggressive antitrust enforcement is often based on a concern that government intervention will be erroneous and societal harm will result from the risk of error. For example, were the Bureau to require the divestiture of assets because of a mistaken belief that the divestiture was necessary to preserve competition, the result might be an inefficient loss of integrated assets. However, it is difficult to envision the societal harm that would result from even an erroneous insistence that senior executives in major firms have a personal incentive to aggressively compete.

In developing this proposal, some questions have been raised about whether the incentives to compete created by the compensation packages we propose will result in too much competition. To the extent that critics suggest that Royal Bank and Toronto Dominion need to have the cushion of non-vigorous rivalry in secure domestic markets in order to utilize the supracompetitive profits from these markets to invest and compete globally, we believe this exposes a defense not heretofore recognized under the Competition Act.⁽¹⁷⁾ It is possible, we concede, that the package might increase the incentive for either bank to engage in an inefficient strategy of raising rival's costs (for example, by refusing to cooperate efficiently in the context of trade associations or the check clearing system for fear that any innovations might benefit its rival disproportionately). Our proposal would focus each dominant firms' incentive on rivalry with the other dominant firm; we emphatically do not suggest that RB and TD management be given bonuses for increased dominance over fringe firms, or put options in small or fringe rivals, precisely because of the risk that it would give them to engage in anti-competitive strategic behaviour.

We readily concede, however, that the attractiveness of the use of either the market share or the financial instrument approach to deter collusion or interdependent behaviour is limited for industries where society prefers a significant amount of inter-firm cooperation through joint ventures and private standard setting. The compensation packages contemplated under this approach would create an increased incentive for managers to forego a pro-competitive and efficient joint project if they believed that the potential benefit to their rival was greater than the benefit to their own company; without this package, the managers would be likely to agree to proposals that increase their own company's net profitability.

We suspect that this concern is overstated. Failure to cooperate in a pro-competitive manner could have sufficiently adverse consequences for the manager's own firm that it does not pose a real risk--TD is not going to withdraw from Interac because the venture might cause a slightly higher rate of profitability for RB, due to its slightly higher market share. Moreover, if the joint venture creates a new product that can be individually marketed by each member, or serves principally to reduce each participant's marginal costs, the venture may increase revenues, sales, and customer base, but will not significantly affect the profitability (and hence the stock prices) or relative market shares of venture participants, who will presumably compete with each other to bid down prices to the level of costs.

Ultimately, whether the nature of pro-competitive cooperation among banking firms is such that our proposed compensation package overdeters socially beneficial cooperation is a factor that the Competition Bureau must assess. The attractiveness of the financial instrument approach in requiring minimal rules and supervision by government authority must be weighed against this potential for over deterrence. Although the market share approach requires the definition of relevant markets and greater care in ensuring the incentives achieve the goal of discouraging interdependent behaviour, it would be less likely to skew incentives for pro-competitive cooperation. A proposal for a joint product that would actually increase TD's market share at the expense of RB is unlikely to receive the assent of RB managers currently, and this proposal would not alter the incentives to agree to joint activity that left market shares unaffected but might marginally increase the profitability of a rival.

Under the Competition Act, proposed mergers whose effect is likely to lessen competition may be blocked by the Competition Tribunal on application of the Competition Bureau's Director. In lieu of blocking the merger or a part thereof, the Tribunal may prohibit the merging parties "from doing any act or thing the prohibition of which the Tribunal determines to be necessary to ensure that the merger or part thereof does not prevent or lessen competition," or may order the merging parties to take any other action, with consent of the parties and the Director.⁽¹⁸⁾ In light of precedents providing that the remedial provisions of the Act are to be read broadly, one might make the argument that the Act permits the Tribunal, under this provision, to order the banks to refrain from compensating their senior executives in a manner that does not provide a strong financial incentive for the executives to seek to increase the banks' profit and/or market share vis-a-vis its chief rival.⁽¹⁹⁾ In this case, this interpretive question may not have to be addressed, for several reasons.

First, bank mergers also require the approval of the Finance Minister, who has broad discretion in the area.⁽²⁰⁾ The Minister could condition his approval upon the banks consenting, per the Competition Act, to the issuance of such an order from the Tribunal, or could require the banks to make undertakings to his ministry that their senior executives will receive pro-competitive compensation packages along the lines we have outlined above.

Second, assuming that the Director will be able to establish that the mergers do lessen competition substantially in one or more specific markets, the banks will recognize that the identification of those efficiencies that merit consideration under s. 96, and the determination of whether those efficiencies "will be greater than, and will offset," the anti-competitive effects of the merger, is a subjective and speculative enterprise.⁽²¹⁾ Rather than risk a Tribunal order blocking the merger, the banks may well agree to these undertakings as a condition of securing the Director's approval. Indeed, the adamant refusal by the banks to agree to a pro-competition compensation should be closely scrutinized by the Bureau and the Tribunal in its analysis of the efficiency defense under s. 96. If analysis suggests that the profitability of these mergers depends on their effect on lessening competition in some Canadian banking markets, in order for Royal Bank and Toronto Dominion to use supra-competitive profits from those markets for other operations, the banks very refusal may provide supplemental inferential evidence that the efficiencies are not greater than, and do not offset the harm to competition.

Finally, and significantly, the proposal may actually provide some benefits to the shareholders of the merging banks (although, presumably, not as many benefits as the ability to obtain monopoly profits from a competition-lessening merger). Corporate boards of directors commonly include stock options as a substantial component of executive compensation, in order to more closely link employees' financial incentives with those of stockholders. These packages are created in the hope that it will spur employees to maximize their own efforts to increase shareholder value. Inevitably, however, stock options can reward employees for increases in the stock that have nothing to do with legitimate employee activities. General improvement in economic conditions can lead to general increases in stock value, even if the performance of the employees is not particularly noteworthy. (Conversely, a general decline in economic conditions can lead to decreases in stock value, even if employees have been doing a marvelous job; managers who would like to be rewarded for their own efforts should prefer the financial instrument approach we propose, and companies should prefer to hire those sorts of managers.) Changes in technology can result in overall increases in industry profitability, again without any particular acumen on the part of current employees. Of particular concern here, mergers that lessen competition and permit the company to reap supra-competitive profits will also increase share values, without (to use a phrase from American antitrust jurisprudence) any superior skill, foresight or industry on the part of management.⁽²²⁾ Precisely because neither shareholders nor directors can accurately disaggregate stock increases to attribute the amount due to management skill, the use of traditional stock compensation packages are necessarily overbroad. Our proposal results in a more precise tailoring of the compensation package. To the extent that managers no longer need to be rewarded because of exogenous economic factors, shareholders are better off. To the extent that a combination of call and put options eliminate industry specific risk beyond the managers' control, the managers are better off. To the extent that managers and shareholders are no longer rewarded because ordinary economic incentives lead management to exploit consumers, the public is better off.

Although it would not be inaccurate to call our proposal unprecedented, we prefer to think of it as innovative.⁽²³⁾ Certainly, it would seem to be as cutting-edge as the notion that new technology for the 21^st Century justifies and requires the consolidation of the Canadian banking industry into a triopoly.

1. The pioneering work in the field was Oliver Williamson, Economies as an Antitrust Defense: The Welfare Tradeoffs, 58 Am. Econ. Rev. 19 (1968). Professor Williamson developed his presentation in a number of articles. See, e.g., Oliver Williamson, Economies as an Antitrust Defense Revisited, 125 U. Pa. L. Rev. 699 (1977). For further discussion concerning these tradeoffs, see Alan A. Fisher & Robert H. Lande, Efficiency Considerations in Merger Enforcement, 71 Calif. L. Rev. 1582 (1983); Paul T. Crampton, The Efficiency Exception for Mergers, 21 Can. Bus. L. J. 371, 378 (1993).

2. To be more precise, the view stated in text is the prevailing American view since acceptance of the Williamsonian trade-off. Prior to that time, American courts often viewed efficiencies as anti-social threats to the competitive process. See Fisher & Lande, supra, note 1, at 1582 & nn. 4-5 (citing cases holding that efficiencies actually weighed against the merger's legality).

3. 374 U.S. 321 (1963).

4. The Competition bureau defines "interdependent behaviour" as "explicit or implicit understandings among firms in the market to jointly exercise market power or limit competition on price, quality, variety, or any other dimension." This type of behaviour is distinct from co-operative behaviour that has the effect of increasing the efficiency with which firms supply their products. Banks have several such co-operative ventures, including the Interac network, and the Bureau recognizes that such ventures can benefit consumers. See Competition Bureau, The Merger Enforcement Guidelines as Applied to a Bank Merger (July 5, 1998) [hereinafter MEG Application to Banks].

5. We confine our analysis to economic exploitation of ordinary Canadian consumers and small businesses. We appreciate that the Finance Ministry may legitimately be considering other social policies such as the cohesion and viability of small communities, unemployment, and political concerns about reliance on foreign banks to provide a substantially increased level of essential banking services.

6. The Bureau's policy statement recognizes that because of the number of products and services supplied by banks, there will be a number of relevant markets where a merger could potentially lessen competition, id. s. 22; that price discrimination, which may result in the finding of narrower markets where market power may be exercised against disfavoured consumers, is facilitated in banking markets because of the exchange of detailed information about bank loans, etc., id. s. 25; that for customers using banks for frequent and small transactions, the geographic market is likely to be smaller, id. ss. 38-39; that small business owners may need frequent personal contact with a loan issuer and the loan issuer may need to closely monitor small business operations, id. s. 41, which is the sort of activity that foreign electronic banking or other potential entrants are unlikely to be able to do; that factors that the Bureau believes will affect the ability of firms to engage in interdependent behaviour, such as transparent fees, stability of underlying costs, frequent transactions, and multi-market exposure, id. s. 68, all seem present in local banking services; and that electronic banking services requiring a computer may not be available to many households and small business, id., s. 75.

We recognize that s. 92(2) of the Competition Act precludes the Tribunal from finding a lessening of competition solely from evidence of concentration or market share. We therefore assume, without deciding, that the factors discussed in the paragraph above (particularly those identified in s. 67 of the MEG Application to Banks) and other aspects of the banking industry would lead the Bureau to find a likelihood of increased interdependent behaviour when the number of Canadian national banks decreased from five to three. But see John Kwoka, Does the Choice of Concentration Really Mater?, 29 J. Indus. Econ. 445 (1981) (existence of three strong rivals is key to competitive industrial performance),

7. This market definition is offered purely for illustration without suggesting that it necessarily aggregates any economically relevant markets.

8. $50,000 x 4 = $200,000.

9. The simplest bonus, for example, would simply be tied to the absolute quantity (of say, checking account balances) produced. Any bonus conditioned on the firm's output instead of its profit would given the manager a larger incentive to price aggressively and resist tacit collusion.

10. Because the recent round of mergers will create two dominant firms, it seems appropriate to confine the undertakings to put options in Royal's principal rival, Toronto Dominion, and vice versa. The Bureau could condition approval of mergers in industries with several remaining firms of roughly equal size by requiring put options in each of the firm's principal rivals.

11. Or equivalently, the manager could buy 5,000 shares in the rival on the open market for $100/share and then immediately turn around and exercise the put thereby selling them for $120/share.

The companies should have some flexibility in how the 10% requirement was met by trading off the number of puts against the degree to which a put was in the money. But because regulators should be interested in preserving the manager's marginal incentives to compete aggressively, it might also be useful to require that the put exercise price be at least 20% (or, alternatively at least one standard deviation above) above the rival's current stock price. If the exercise price were set too closely to the current price, managers will simply ignore their puts and will again have an incentive to simply increase their own company's stock price through interdependent behaviour. For example, consider the result if the exercise price for a put on TD stock was $101; once it was clear that, due to general economic conditions in Canada, TD stock had risen to $105, the value of the put would be lost, absent some extremely aggressive move by RB that would actually drive down the stock price. Once the value of the put is lost, managers no longer have an incentive not to collude or engage in interdependent behaviour.

12. The exercise prices of the puts and calls cap the manager's incentive to a certain range of stock prices. In the textual example, the RB manager is indifferent between year-ending RB share values of $70 and $75 because her calls have an exercise price of $80 (similarly for TD year-end prices above the $120 put exercise price). It is, however, possible for RB to enter into a compensation contract that does not have this capped attribute. The contract in a sense would emulate the payoffs the manager would have if she bought (long) shares of RB and sold (short) shares of TD. Such a contract might have the following bonus/penalty formula:

B = 5000(RB-80) + 5000(120-TD),

where RB and TD are the year-end stock prices of the two respective banks. If "B" is positive, this bonus would be added to the $800,000 base pay. If "B" ends up being negative, this penalty would be deducted from the base salary. The manager's total compensation can be expressed more simply as:

$1,000,000 + 5,000(RB-TD).

13. United States v. Aluminum Co. of America, 148 F.2d 416, 427 (2d Cir. 1945) (Hand, J.).

14. Ed Miller Sales & Rentals Ltd. v. Caterpillar Tractor Co.. 69 C.P.R. 3d 143, 156 (Alta. C.A. 1996).

15. Canada (Director of Investigation & Research) v. Southam, Inc., 144 D.L.R. (4^th) 1, 28 (S.C.C. 1997).

16. Ibid. at 30.

17. The Act specifies that a "redistribution of income between two or more persons" does not constitute an efficiency. Competition Act, s. 96(3).

18. Competition Act, s. 92(1)(f)(iii).

19. Section 12 of the Interpretation Act, R.S.C. 1985, c. I-21, provides that statutes "shall be given such, fair, large and liberal construction and interpretation as best ensures the attainment of its objects." The Competition Act's express purpose is to "maintain and encourage competition in Canada". Competition Act s. 1.1. In American Airlines Inc. v. Canada (Competition Tribunal) (1988), 54 D.L.R. (4^th) 741, 748, 23 C.P.R. (3d) 178, 185, [1989] 2 F.C. 88 (per Iacobucci C.J.), the court referred to this purpose as "extremely broad," and i that case and in the Chrysler Canada Ltd. v. Canada (Competition Tribunal) (1992), 92 D.L.R. (4^th) 609, 42 C.P.R. (3d) 353, [1992] 2 S.C.R. 394, both courts gave a broad interpretation to the companion legislation, the Competition Tribunal Act. Moreover, in Bell Canada v. Canadian Radio-television and Telecommunications Commission (1989), 60 D.L.R. (4^th) 682, 707, [1989] 1 S.C.R. 1722, Gonthier, J., emphasized that courts must "avoid sterilizing these powers [of regulatory authorities] through overly technical interpretations of enabling statutes."

It is true that, in the context of completed mergers, the courts have held that the Tribunal's options are limited to orders dissolving the merger or ordering divestiture. See Director of Investigation & Research v. Air Canada, 104 D.L.R. (4^th) 129 (Fed. Ct. Appl. 1993). However, the relevant section cited in text empowering the Commission to issue additional orders against the merging parties applies only to proposed mergers. But see id. at 144 (opinion of Hugessen, J.A.) (Parliament intended to limit Tribunal's powers to "blunt instruments" in order to facilitate negotiated settlements between the Director and the merging parties).

20. Section 255(2) of the Bank Act permits the Finance Minister, "with approval of the Governor in Counsel, but otherwise at his sole discretion," to issue letters patent amalgamating two or more banks into one. "(The French translation is clearer, providing that "le ministre peut ... delivrer, a sa discretion, des lettres patentes les fusionnant en une seule personne morale constituant une banque unique.")

In addition, the approval of the two pending bank mergers may be considered by the Government as part of a major initiative to deregulate financial services. Assuming this initiative took the form of new legislation, conditions on the two mergers could be incorporated into that legislation.

21. The Competition Bureau's guidelines concerning bank mergers note that an otherwise anticompetitive merger can only be saved if the demonstrated efficiencies are both "more weighty than, more extensive than, or of a larger magnitude than the anticompetitive effects" and that the efficiency gains "must neutralize, counterbalance or compensate" for the likely anticompetitive effects of the merger. See MEG Application to Banks, supra note 4, s. 107. Where, as is likely the case here, efficiency gains and anticompetitive effects cannot be weighed in similar terms, the evaluation is necessarily "subjective in nature and will ordinarily require the exercise of the Director's discretion." Id. s. 108.

The scope of the Director's discretion is enhanced because the guidelines interpret the statutory efficiency defense contained in s. 96 of the Competition Act in a significantly more favorable manner for the merging parties than the interpretation that might be given by the Tribunal or the courts. The Bureau has stated that, in weighing the efficiency claims against anticompetitive harms, it will only consider the "deadweight loss to the Canadian economy" caused by the reduced output likely to occur from a competition-lessening merger. Id. s. 109. At least one court, in dicta, has suggested that efficiency gains must also offset harms caused by the transfer of wealth from Canadian consumers to the banks. See Director of Investigation & Research v. Hillsdown Holding (Canada) Ltd. [1992] 41 C.P.R. 3d 289. For a suggestion that the Bureau's position does not represent the best interpretation of s. 96, see Stephen F. Ross, Did the Canadian Parliament Really Permit Mergers that Exploit Canadian Consumers so the World Can Be More Efficient?, 65 Antitrust L.J. 641 91997).

Perhaps the Bureau Director would consider revising his policy to conform his legal interpretation with that of the Hillsdown dicta while adhering to the total welfare standard as a matter of enforcement discretion. A reading of the Competition Tribunal's remedial powers to preclude a curial requirement of pro-competitive compensation for leading bank executives, see note 17 supra and accompanying text, renders even more problematic the Bureau's policy of ignoring a competition-lessening merger's distributional effects. This narrow reading would render the Bureau powerless to seek a remedy that could save Canadian consumers millions of dollars without any loss of efficiency because the Director could not invoke the Tribunal's powers, as a threshold matter, where efficiency gains exceeded deadweight loss.

22. Cf. Alcoa, 148, F.2d at 430.

23. Others have suggested, in related fashion, that firms not be permitted to invest in their rivals' stock because this tends to facilitate tacit collusion. See David Gilo, The Anticompetitive Effect of Passive Investment, John M. Olin Discussion Paper No. 189, Harvard Law School (June, 1998).