Executive Summary

Introduction

Meridian was asked by a group of six Canadian banks (“the Banks”) to review horizontal benchmarking and its potential impact on CEO Compensation and pay disparity to assist with their consideration of concerns raised in shareholder proposals. In particular, our research focused on the causes of increasing CEO pay, possible risks associated with horizontal benchmarking and potential safeguards and alternative approaches.

Our research included a review of articles provided by Northwest and Ethical Investments (“NEI”) as well as additional academic articles focused on executive pay and the use of horizontal benchmarking. Our review included a cross section of materials and reflected a spectrum of viewpoints. Our research focused primarily on the U.S. and Canada; a majority of the reviewed articles are from the U.S., as the issue has received significantly more academic attention in the U.S. than Canada. The U.S. has also had some form of executive pay disclosure since the 1930’s (Canada only since the 1990’s), so long term pay comparisons typically rely on U.S. information.

In addition to academic studies, we have analyzed compensation trends and practices among the Banks and the broader market based on data provided by the Banks and available through public filings. Our research focused on trends in compensation for top executives, as well as several other positions at different levels within the Banks. We also reviewed data on executive turnover provided by the Banks.

Trends in CEO Compensation Levels

There appears to be general acceptance that executive pay has:

  1. Increased significantly over the last 40 years; and
  2. Increased at a faster rate than median employee compensation.

However, more recent trends show a significant downward movement in executive compensation and executive compensation relative to median employee compensation since 2000.

Based on our review of compensation at the Banks, CEO compensation reflects the broader market results. Bank CEO compensation increased significantly in the late 1990’s, with the increase predominantly in long term incentive compensation. Total Bank CEO compensation peaked in 2001, but since then has declined marginally in real terms. Total cash compensation for Bank CEO’s has declined 26% in real terms since 2000.

While Bank CEO total direct compensation has remained relatively flat since 2000, broader Bank employee compensation has increased. As a result, the ratio of Bank CEO pay to employee pay has declined since 2000. The total cash compensation ratio declined from 66:1 in 2000 to 45:1 in 2012 and the total direct compensation ratio declined from 145:1 in 2000 to 135:1 in 2012.

This downward shift in compensation and pay ratios is relatively recent and it remains to be seen whether it will be a continuing trend or is a shorter-term result of recent economic circumstances in combination with better pay for performance alignment.

Causes of High Executive Pay

There are a myriad of conflicting theories regarding the increase in executive compensation and no consensus or single compelling reason for high executive pay. In particular, none of the articles we reviewed identifies a strong causal connection between pay levels and horizontal benchmarking practices, rather horizontal benchmarking is one of many practices which are correlated with high executive pay, but for which a causal connection is not established. However, certain problematic benchmarking practices, such as targeting compensation above median for target performance and benchmarking compensation relative to an “aspirational” (i.e., significantly larger) comparator group appear likely to result in increases in executive compensation.

Other suggested causes of high executive pay are as follows:

  • Increased disclosure requirements have driven increases in executive pay by making comparative pay information more readily available. Arguably, the disclosure of executive pay is more closely correlated with increases in pay than benchmarking.
  • The tax and accounting treatment of share-based compensation has, over different time periods, promoted the use of options and impeded good governance practices, thus resulting in higher pay particularly through periods of market growth, although with increased alignment of pay with shareholder experience.
  • Changes to government policies (in particular, policies which were implemented to reduce executive compensation such as U.S. excess parachute taxes) have often had the unintended effect of significantly increasing executive pay.
  • Executive pay levels have correlated directly with the increased size and complexity of businesses. As CEO responsibilities change more directly with increased size and complexity of the business, this may also explain the increased disparity between CEO and employee pay.
  • The shift in CEO skills from company specific to general management skills evidenced by mid and late career transition and changes in executive education has made CEOs more mobile, requiring higher and more competitive pay for attraction and retention.
  • The increased focus on pay for performance has resulted in higher compensation at risk and increased CEO turnover, which has resulted in higher pay levels to compensate for the increased risk.

There is no clear single cause of high executive pay and horizontal benchmarking has been, at most, only one factor associated with increased executive pay levels.

Alternatives to Horizontal Benchmarking

Various methods of vertical benchmarking, or comparing CEO compensation to that of other employees, have been discussed as possible alternatives to horizontal benchmarking. Vertical benchmarking ratios between the CEO and all employees have received the most attention, particularly given the requirement (which has not yet been implemented) in the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (“Dodd-Frank Act”) requiring companies to disclose the ratio between CEO pay and the median of all other employees.

A few companies have attempted to establish a cap on CEO pay based on a multiple of compensation provided to the broader workforce. However, these examples provide evidence of the challenge of attracting and retaining executive talent within such a constraint. These companies have used ratios that exclude equity compensation, which serves as a significant component of executive pay, and have increased the maximum multiple or abandoned the practice in order to attract and retain executives.

None of the articles we reviewed suggested that benchmarking should be completely eliminated and even those writers who attributed significant increases in executive pay to horizontal benchmarking acknowledged that a human resources and compensation committee (“Committee”) needs some basis on which to set the upper and lower parameters of executive pay. Writers who recommended vertical benchmarking (assessing executive compensation as a multiple of regular worker pay) typically suggested that it can provide important context but is not sufficient to serve as a stand-alone solution.

While vertical benchmarking is unlikely to be sufficient as a primary basis for setting executive compensation, it can provide important context for a Committee, particularly in assessing trends in pay disparity.

Safeguards to Ensure Appropriate Use of Horizontal Benchmarking

Horizontal benchmarking provides important context for Committees when setting executive compensation levels. Several safeguards can assist Committees to ensure that horizontal benchmarking does not lead to inappropriate escalations of executive pay. These safeguards include development of a size and industry appropriate comparator group, thoughtful application of horizontal benchmarking data taking business context into account, and following effective corporate governance processes in compensation decision making.

Conclusion

The causes of increasing executive pay are complex and most likely based on many factors, rather than a single factor. There is no compelling evidence to conclude that horizontal benchmarking results in excessive executive pay. However, horizontal benchmarking, used inappropriately and without safeguards, can be used to support inappropriately high pay levels. When horizontal benchmarking is used properly it provides relevant context and is an important input for Committees when setting executive pay and is a key input in setting pay parameters. While vertical benchmarking is unlikely to be sufficient on its own to guide Committees establishing executive compensation, it can provide additional context to assist decision making.

Introduction

Meridian was retained by the Banks to assess the use of horizontal benchmarking and its potential impact on CEO compensation and pay disparity. Specifically, Meridian was asked to:

  • Summarize the core areas of concern with horizontal pay benchmarking and the related risks to the Banks, including a review of recent academic articles with a focus on evidence supporting the position that horizontal benchmarking causes excessive CEO compensation;
  • Identify potential alternatives to horizontal pay benchmarking and supplemental analyses to augment current practices, with the pros and cons of each of these approaches and examples of firms employing these approaches; and
  • Summarize trends and anticipated changes in the approach to pay benchmarking and the establishment of executive pay levels, including the potential impact of the Dodd Frank pay ratio disclosure.

The Banks retained Meridian for this project as part of their collective commitment to examine issues raised by NEI and other shareholders. NEI and the other shareholders expressed concern that the use of horizontal benchmarking is increasing the pay disparity between executives and other employees, potentially creating risks for the Banks by lowering the engagement and motivation of lower-level employees. The shareholders asked the Banks to consider alternatives to horizontal benchmarking, including vertical pay ratios.

Our research included a review of articles provided by NEI, as well as additional academic articles focused on executive pay and the use of horizontal benchmarking. Our review did not include articles written by compensation consultants, as they often facilitate the benchmarking process and have been viewed by some critics as contributing to excessive executive compensation. Additionally, our research focused primarily on the U.S. and Canada; a majority of the reviewed articles and data are from the U.S., as the issue has received significantly more academic attention in the U.S. than Canada. The U.S. has also had some form of executive pay disclosure since the 1930s (Canada only since the 1990’s), so long term pay comparisons typically rely on U.S. information.

In addition to academic studies, we have analyzed compensation trends and practices among the Banks and more broadly, based on data provided by the Banks and available through public filings. Our research focused on trends in compensation for top executives as well as several other positions at different levels within the Banks. We also reviewed data on executive turnover provided by the Banks.

Meridian was asked by the Banks to provide a full perspective on the issues included in this review and, in particular, to identify concerns with horizontal benchmarking so that each of the Banks could evaluate their current practices for setting executive pay and consider potential modifications or alternatives. The Banks did not suggest or request any particular outcome or conclusion, other than a thorough review of the issue and suggestions for improving executive compensation setting practices. This report summarizes the complete findings of our review.

Key Articles

Key articles reviewed are listed at the end of this report and copies are included at Tab A of the report. Our review of articles was not exhaustive, as there are hundreds of articles on this topic. We focused on a cross section of articles which provide different perspectives in order to canvas a range of viewpoints.

Is CEO Pay Increasing on an Inappropriate Basis?

Executive pay increased significantly in the latter part of the 20th century, and the increase was disproportionate to the increase in compensation of ordinary employees. Studies suggest this trend has reversed since 2000. This downward shift in compensation is relatively recent and it remains to be seen whether it will be a continuing trend or is a shorter-term result of recent economic circumstances in combination with better pay for performance alignment.

Is CEO Pay Increasing?

U.S. CEO pay increased over 400% in the 1990s2. However, since 2000, CEO pay has decreased 40% in real terms.3 In that same period, CEO compensation has become more variable and there is a greater difference between average and median compensation.4 Average CEO pay relative to operating income has returned to the levels of the mid-1990s, and relative to net income is lower than it was in the mid-1990s. Over the same 10-year period, CEO turnover has increased somewhat.5 Over the last 20 years, CEO pay relative to the top 0.1% of the population has remained relatively constant,6 and is the same currently as in 1994 for S&P 500 CEOs and is lower than 1994 for non-S&P 500 CEOs. Other high earners such as athletes and lawyers have also seen significant increases in pay, at least as large as those of CEOs7, suggesting that CEO pay has risen for reasons unrelated to benchmarking or the “agency” model of corporate law.

Murphy, who has reviewed more recent trends in CEO pay, summarizes this recent shift in CEO pay as follows:

The first decade of the new century have (sic) brought several important changes in the level and composition of CEO pay…. median grant-date total CEO pay in the S&P 500 declined from $9.3 million in the peak of the year of 2001 to $9.0 million in 2011, representing the first prolonged stagnation in CEO pay…Workers have actually done better than CEOs….After accounting for inflation, average-worker pay has risen by 4 percent since 2000, while CEO pay has fallen by 30 percent. Compared with the levels right before the recession started, average-worker pay, adjusting for inflation, has gone up by 2.2 percent while CEO pay has risen by 5 percent.

The ratio of CEO pay to worker compensation has also declined since 2000, which was the peak of the disparity. One of the articles provided by NEI notes the significant decline in the pay disparity since 2000 as follows:

Using a measure of CEO compensation that included the value of stock options granted to an executive, the CEO-to-worker compensation ratio was 18.3-to-1 in 1965, peaked at 411.3-to-1 in 2000, and sits at 209.4-to-1 in 2011. Using an alternative measure of CEO compensation that included the value of stock options exercised in a given year, CEOs earned 20.1 times more than typical workers in 1965, 383.4 times more in 2000, and 231.0 times more in 2011.

Is the Disparity Between CEO Pay and Other Employee Pay Increasing at the Banks?

Meridian reviewed compensation data provided by the Banks to assess trends in CEO pay and the ratio between CEO pay and that of the broader employee population. Each bank provided at least 18 years of CEO compensation data (from 1995 to 2012). Historical compensation data for other employees, however, was more limited. Based on available data, Meridian compiled an “Employee Composite” that includes data beginning in 2000 for several jobs from the broader employee population. Roles represented in this composite include Customer Service Representative, Customer Service Manager, Account Manager, Personal Banker, Sales Manager, and Retail District Manager.

The change in the Banks’ CEO pay data over the past 18 years mirrors the broader market results discussed above. As shown in the chart below, CEO compensation increased significantly in the late 1990s, with the increase predominately seen in long-term incentive grants. Total compensation peaked in 2001, but has since declined marginally in real terms. Total cash compensation has decreased 26% in real terms since peaking in 2000.

While CEO total direct compensation has remained flat since 2000, broader employee compensation has increased. The chart below compares the trends in pay for the CEO and the Employee Composite. CEO pay was much more volatile than the Employee Composite, reflecting the increased amount of pay at risk for CEOs.

Consistent with the broader market findings discussed above, the ratio of CEO pay to the Employee Composite pay has declined since 2000. The total cash compensation ratio declined from 66:1 in 2000 to 45:1 in 2012. Likewise, the total direct compensation ratio declined from 145:1 in 2000 to 135:1 in 2012.

What is Horizontal Benchmarking?

Most companies use horizontal benchmarking (either formally or informally) as one important input in setting executive compensation. Horizontal benchmarking involves comparing compensation for an executive role at one company to compensation for similar roles at a group of peer companies to assess market competitiveness of pay levels, pay mix, pay design, as well as other compensation governance factors such as share ownership and termination and change in control provisions.

The first step in a competitive benchmarking of executive compensation is generally to develop an appropriate comparator group considering:

  • Size of the business. This is frequently used as a proxy for the complexity of an organization. The primary measure used to assess size should be appropriate to the business, (i.e., revenue will generally be used for manufacturing and retail businesses, assets will generally be used for financial services, mining, and oil and gas businesses, and market capitalization or enterprise value will generally be used for startups). Size parameters are commonly established (e.g., only companies from ½ to 2 times the size of the subject company will be included) and typically the goal is to position the subject company close to the median of the comparator group, as size is closely correlated with level of executive pay. If a group of appropriately sized similar companies is not available, compensation data can be regressed to provide size-adjusted competitive information.
  • Industry. Similar types of businesses provide for better comparison of the scope and complexity of executive roles, more applicable information respecting design of pay programs (including choice of performance metrics) and similar challenges and complexities.
  • Competitors for talent. Comparator groups typically consider which companies the subject company looks to when it is recruiting, as well as to which companies it loses talent. Where there is a large disparity in size between the subject company and a competitor for talent, that competitor’s pay practices may be tracked for “interest”, rather than included in the comparator group for benchmarking, or pay may be regressed to provide size appropriate information.
  • Geographical area of operations. Generally companies choose, as their comparators, companies which operate in the same jurisdictions as they operate. Including companies from other jurisdictions is supportable when it reflects the company’s competition for talent, jurisdiction of operation or when there are few or no relevant comparators in the company’s home jurisdiction. This is particularly an issue for Canadian companies with North American operations, as U.S. pay practices can differ significantly from Canadian practices in terms of level, design and leverage.

The second step in benchmarking compensation is to determine appropriate position matches. A CEO is generally matched to other CEOs, but matches for other positions can be more difficult, because of the different titles companies use for similar positions and because some roles may be significantly different as a result of the company’s organizational structure. CFO benchmarking can be challenging despite the required disclosure in the proxy circular, as there can be large disparity in the scope of the role (i.e., some CFOs have a financial role only, while others also have a significant strategic role) and pay level reflects the scope of the role.

The third step is to compare pay levels, pay mix and plan design of the subject company to the comparator companies and to identify the competitive market position of the subject company’s pay practices.

Horizontal benchmarking is used extensively below the executive level through the use of compensation surveys to set compensation for technical, sales, administrative and other talent.

Horizontal benchmarking is also a standard business management tool in other contexts as well, such as review of performance relative to peers or industry and governmental benchmarks. These relative measures are a form of horizontal benchmarking which is extensively used by investors and other stakeholders. For example, a company may set occupational safety targets based on comparator company levels of injury and accident or government benchmarks.

Horizontal benchmarking provides an important perspective on the external market for talent.

Does Horizontal Benchmarking Cause or Contribute to Excessive Executive Pay?

NEI Material

Most of the articles provided by NEI do not directly address whether horizontal benchmarking causes excessive executive pay; instead, they focus primarily on a number of important social concerns, including:

  • That executive pay is rising at a disproportionate rate to median employee compensation
  • The societal benefits of responsible environmental, social and governance practices
  • The potential social destabilizing and negative broad economic effects of significant income inequality
  • Reduced productivity in workplaces where there is a greater disparity between CEO and regular employee annual incentive pay
  • The increased turnover of general employee populations if executive compensation is too high
  • The damage that may be caused by shareholder primacy and “short termism”
  • A lack of correlation between high incentives and high performance
  • The potential increase in fraud or “gaming” when incentives are too high
  • That incomes of the super-rich are now more closely tied to compensation than to property ownership

While these are all important societal issues, many are focused more broadly on the impact of high executive pay and income disparity as opposed to the practice of horizontal benchmarking. We did not review all of these issues or the studies suggesting that increased incentives do not correlate with improved performance as they are outside the scope of our review and our expertise. We focused instead on whether horizontal benchmarking is a cause of excessive executive pay.

Assessing the Relationship between Horizontal Benchmarking and Excessive CEO Pay

Among writers who do assert that horizontal benchmarking causes excessive executive pay are Elson and Ferrere, who comment:

In setting the pay of their CEOs, boards invariably reference the pay of the executives at other enterprises in similar industries and of similar size and complexity. In what is described as “competitive benchmarking,” compensation levels are generally targeted to either the 50th, 75th, or 90th percentile. This process is alleged to provide an effective gauge of “market wages” which are necessary for executive retention. As we will describe, this conception of such a market was created purely by happenstance and based upon flawed assumptions, particularly the easy transferability of executive talent. Because of its uniform application across companies, the effects of structural flaws in its design significantly affect the level of executive compensation.

Elson and Ferrere assert that the practice of horizontal benchmarking is significantly responsible for excessive executive pay:

(I) theories of optimal market-based contracting are misguided in that they are predicated upon the chimerical notion of vigorous and competitive markets for transferable executive talent; (II) that even boards comprised of only the most faithful fiduciaries of shareholder interests will fail to reach an agreeable resolution to the compensation conundrum because of the unfounded reliance on the structurally malignant and unnecessary process of peer benchmarking; and, (III) that the solution lies in avoiding the mechanistic and arbitrary application of peer group data in arriving at executive compensation levels. Instead, independent and shareholder-conscious compensation committees must develop internally created standards of pay based on the individual nature of the organization concerned, its particular competitive environment and its internal
dynamics.

They recommend that companies set executive pay based on internal pay equity. However, they note that:

Admittedly, our prescription is not concrete or easily implemented, but as the shareholder value movement has empowered directors as never before to act in their investors’ interest, any solution to the compensation conundrum must be founded upon their expert judgment and discretion.

We agree with Elson and Ferrere that a collective practice of targeting compensation above the median of the comparator group inevitably leads to an upward movement of executive compensation.

Elson and Ferrere note that board governance has significantly improved since the early 1990’s, and thus they do not agree that high levels of executive pay result from ineffective governance (weak boards). However, the governance improvements which started in the 1990’s were implemented gradually and appear to be more aligned with the 40% decrease in executive pay since 2000. This suggests that improvements in governance practices may, in fact, be having a salutary effect on executive compensation levels.

Conclusion

Elson and Ferrere do not provide a basis to conclude that horizontal benchmarking inherently causes excessive executive pay. However, they do provide useful commentary on some of the poor practices sometimes associated with horizontal benchmarking and that can contribute to inappropriate increases in executive pay (which are discussed in more detail in the following section). They did not comment on the decline in executive pay over the last 12 years, during which time improved standards for horizontal benchmarking have emerged. When considered in combination with the numerous other factors which have caused or contributed to increasing executive pay (see the discussion under the heading “Are There Other Significant Influences on Executive Pay?”), it appears that horizontal benchmarking is, at most, only a minor factor associated with increased executive pay levels.

What Are the Significant Concerns With Horizontal Benchmarking?

While there is no strong correlation between horizontal benchmarking and increased or excessive executive pay, two practices associated with horizontal benchmarking, targeting pay above median and using an aspirational peer group, by their nature, are likely to contribute to increases in pay levels.

Targeting Pay Above Median

The practice of targeting compensation at levels above the peer group median is widely believed to have a “ratcheting-up” effect on pay.

Though the pay of other executives may provide a beneficial indication of the outside compensation opportunities available and the necessary retention wages, the external references can have perverse systemic effects on the aggregate level of compensation where even only a few executives are overpaid. Thomas DiPrete and Gregory Eirich argue that excessive pay increases for even a relatively small proportion of CEOs can have a significant effect on the pay of the remaining executives. This “leapfrog effect” propagates through the networks constructed by the peer grouping process. As a result, one firm’s overpayment affects all the connected firms for which they are a peer.

Whether the excess compensation is awarded for merit or otherwise, a talented individual who is paid on a scale deserving of their abilities should not, through the peer group mechanism, be allowed to bolster the pay of less able executives; particularly when those executives would be paid less if serving in a similar fashion. An individual executive may deliver phenomenal performance and be likewise compensated, but why should that individual’s contribution to their company be relevant to another individual’s actions at another company. Pay and opportunity wages should reflect and adjust to this evaluation of an individual merit.”

While this comment is generally correct, it fails to take into account two fundamental principles of current benchmarking practices:

  • Review of pay levels is typically based on median data, rather than average, so outliers (those who pay very high or very low), have a much less significant effect on compensation recommendations.
  • Current practices for executive benchmarking are increasingly focusing on target pay, rather than actual pay, which eliminates some of the bias that high performers would otherwise have.

To determine the prevalence with which Canadian companies target compensation above the median, Meridian sampled the Standard & Poor’s/Toronto Stock Exchange 60 (“TSX 60”). If disclosed, target pay levels were captured from each TSX 60 company’s most recently filed proxy statement. Though some companies set targets for total compensation, others target individual elements of pay or do not set targets at all. Thus, for total compensation and for each element of pay, Meridian computed the percentage of TSX 60 companies that target pay at the 25th percentile or below, between the 25th percentile and the median, at the median, between the median and the 75th percentile and at or above the 75th percentile. The chart below depicts the results of this study.

Compensation Targets of TSX 60 Companies Relative to Self-Selected Peer Groups

Element of Compensation Disclosed Target Compensation Positioning Relative to Peer Group
≤ 25th 25th – 50th 50th 50th – 75th
Base Salary 3% 3% 68% 21% 6%
Short Term Incentive 0% 0% 76% 18% 6%
Total Cash 0% 0% 80% 13% 7%
Long Term Incentive 0% 5% 68% 21% 5%
Total Compensation 0% 0% 67% 26% 7%

The results of the study showed that, for total compensation and for each element of pay, the majority of companies target the median level. Few TSX 60 companies target pay below the median. However, a number of companies target pay above the median—33% of the disclosing companies for total compensation and between 20% and 27% of the disclosing companies for specific elements of pay.

In an effort to determine whether U.S. companies target compensation at levels similar to Canadian companies, Meridian followed the same methodology using Standard & Poor’s 100 Index (“S&P 100”) companies as the sample population. The results of the S&P 100 study are set out in the chart below.

Compensation Targets of S&P Companies Relative to Self-Selected Peer Groups

Element of Compensation Disclosed Target Compensation Positioning Relative to Peer Group
≤ 25th 25th – 50th 50th 50th – 75th
Base Salary 0% 0% 85% 9% 6%
Short Term Incentive 0% 0% 86% 14% 0%
Total Cash 0% 0% 60% 27% 13%
Long Term Incentive 0% 0% 76% 16% 8%
Total Compensation 0% 0% 65% 28% 7%

Unlike the TSX 60 companies, none of the S&P 100 companies target total compensation or any element of compensation at levels below median. Similar to TSX 60 companies, approximately one-third of the S&P 100 companies that disclosed total compensation targets (35%) selected levels above the median. For S&P 100 companies, there is more variation by specific pay element in the percentage of companies that target compensation above median (between 14% and 40% of the disclosing companies).

In recent years companies are finding increasingly difficult to support the practice of targeting executive pay above the median. Say on Pay and increased scrutiny from institutional shareholders and proxy advisory firms has led many companies to reevaluate the practice of targeting above-median pay for the achievement of target performance. Generally, targeting pay above median pay is viewed as appropriate only in very limited circumstances (e.g., when a majority of a company’s peers are smaller and the targeted pay positioning aligns with the company’s size relative to the peers).

Meridian also evaluated the targeted pay levels of the Banks, both in 2009 and in 2013.

  • In 2009, all six of the Banks were already targeting either total compensation (or components of compensation) at the peer group median.
  • According to 2013 proxy disclosures, only one of the six banks targeted total compensation above the median. This Bank is larger than its peers which likely accounts for its target pay philosophy.

Thus, in summary, the majority of the six largest Canadian banks have targeted median compensation levels for at least the past five years, with the only exception being reasonable due to the size of the companies in its comparator group.

Using “Aspirational” Peer Groups

Executive pay levels are closely tied to company size. Accordingly, choosing an aspirational peer group can have a similar effect to targeting pay above median. Concerns have been raised that when making comparisons between themselves and others, individuals and companies tend to compare themselves to those who are seen as slightly better or more expert, thus creating an upward bias.

In 2006, the U.S. securities laws were changed to require disclosure of compensation comparator groups. Say on Pay has also led to increased focus on peer groups, and many companies have altered their peer groups to be more in line with their current size based on investor feedback.

In order to determine whether Canadian companies are developing size-appropriate peer groups, Meridian selected the companies within the TSX 60 as a sample population. For each TSX 60 company, Meridian determined the percentile rank (based on sales, assets and market capitalization) as compared to that company’s most-recent proxy-disclosed benchmarking peer group. Of the 60 companies within the TSX 60, only 47 companies currently have benchmarking peer groups that are fully comprised of public companies for which size data is available. Therefore, Meridian’s sample population was a 47-company subset of the TSX 60. When the respective percentile ranks for each of the 47 companies were totaled, both the average and the median percentile rank for each metric (sales, assets and market capitalization) was at a level of 50% +/-6.5%. Thus, TSX 60 companies appear to be selecting benchmarking peer groups for which they are appropriately positioned near the median in sales, assets and market capitalization.

Meridian similarly reviewed the peer groups of the Banks. Unlike the broader TSX 60 study, the Banks commonly compare themselves to smaller peers. On average, approximately two-thirds of the Banks’ self- selected peer groups have assets and market capitalizations smaller than their own. The difference in results may be due primarily to the smaller sample size of the Banks subset and the limited number of comparably sized industry peers.

None of the Banks uses an aspirational peer group for compensation benchmarking. Instead, the Banks are generally using peer groups with a majority of companies that are smaller than they are.

Are There Other Significant Influences on Executive Pay?

There are a number of different and often conflicting studies on the causes of increase in executive pay. We have summarized the most frequently cited theories and identified some of these conflicting views below.

The causes of increasing executive pay are complex and most likely based on a combination of many factors, rather than a single factor.

Increased Disclosure

Based on U.S. and Canadian material, there appears to be a strong correlation between increased disclosure and increased executive pay.

The first required disclosure of executive compensation in the United States was pursuant to the Securities Act of 1934 and was introduced in December 1934, and effective June 1935. Prior to that date, there is little information on U.S. CEO pay. By contrast, in Canada, disclosure of executive compensation was not required until 199426. Gelinas and Baillargeon concluded that:

Following 1993, when mandatory disclosure of CEO compensation was implemented by the Canadian Securities Administrators, compensation started to increase at a greater pace.

In an analysis of Canadian CEO compensation between 1992 and 1997, Gelinas noted that:

…mandatory disclosure forces many firms to align their executives’ pay with the market going rate. This is good news for investors because it achieves the regulation’s goal: firms paying highly restrain pay increases and save money, and firms paying below market limit their chances of incurring costs associated with such a low position by adjusting pay upward… however, mandatory disclosure enables some firms to race for the market’s top pay position. The empirical results show that this overbid for the top paying spot has a market-wide inflationary trend by awarding CEO pay increases that are less linked to corporate performance.

Thus, Gelinas and Baillargeon believe that there is a direct correlation between executive pay levels and the amount of pay information that is available.

Changes to Government Policies, Tax and Accounting Rules

Many writers comment that government policies, tax and accounting rules play a significant role in executive compensation. Murphy comments that:

…government intervention into executive compensation—largely ignored by researchers—has been both a response to and a major driver of time trends in CEO pay. There have been two broad patterns for government intervention into CEO pay. The first pattern is aptly described as knee-jerk reactions to isolated perceived abuses in pay, leading to ‘one-size-fits-all’ responses and a host of unintended and undesirable consequences. The second pattern—best described as ‘populist’ or ‘class warfare’—arises in situations where CEOs (and other top executives) are perceived to be getting richer when lower-level workers are suffering. Beyond these two broad patterns, indirect intervention in the form of accounting rules, securities laws, broad tax policies and listing requirements have also had direct impact on the level and composition of CEO pay. In most cases, companies and their executives have responded to the interventions by circumventing or adapting to the reforms, usually in ways that increased pay levels and produced other unintended (and typically unproductive) consequences.

The inclusion of stock options as a part of executive pay is strongly correlated with increases in executive pay. The prevalence of options was driven in large part by the favourable accounting treatment of options. However, performance-contingent options were not awarded the same favourable treatment, therefore penalizing the use of this more performance focused form of equity.

In 1991, changes to the U.S. securities rules encouraged the practice of selling shares as soon as options were exercised. Before that, many options had post-exercise hold periods. Similar changes in Canadian tax laws, which require that option gains be recognized and taxed at the time of exercise, also discourage the good governance practice of requiring executives to hold the shares received on the exercise of an option for a period of time following exercise. In 2005, the U.S. accounting rules were changed to level the playing field between options and restricted stock and to eliminate the disincentive for performance-contingent options32, resulting in a significant shift in U.S. compensation mix to reduce the proportion of options and increase restricted stock. In Canada, tax laws have not yet caught up with good governance practices, particularly for equity compensation, and options remain a significant component of long term compensation for Canadian executives.

Other examples of government attempts to limit executive pay that, in fact, had the opposite effect, include U.S. attempts to tax golden parachutes and limit the tax deductibility of non-performance based pay. The excess parachute tax in the U.S. (a tax designed to cap severance payments by imposing a surtax on and limiting deductibility of excessive payments in connection with a change of control) attempted to limit change- in-control severance to 3 times total compensation. Instead, the 3 times multiple (which was the threshold for the surtax and loss of deductibility) became the floor for severance and many companies added tax gross- ups to make executives whole for the surtax (significantly increasing the non-deductible cost of these arrangements), vesting periods for options were shortened (to reduce the amount which was included in the calculation) and severance on a change of control was increased.

Likewise, the introduction of IRC Section 162(m) in the U.S., which was designed to penalize non- performance-based compensation, resulted in an increase the prevalence of options, an increase in salaries and a replacement of reasonable discretionary plans with generous formulaic plans.

Increase in Company Size and Business Complexity

Gabaix and Landier correlate an increase in CEO pay of 500% from 1980 to 2000 with a 500% increase in company size. Murphy notes that:

The average market capitalization of firms in the S&P 500 grew (in 2011-constant dollars) from $10.0 billion in 1992 to $35.8 billion in 2000 (before falling to $22.7 billion in 2011).

Frydman and Jenter similarly comment that the increase in CEO pay may be contributed to by the increase in the size and complexity of businesses and the change in firm characteristics and technology which have increased the effect a CEO can have on a business.

This increase in company size and complexity may also explain some of the growth in pay disparity between CEOs and other employees. The CEO has responsibility for managing the whole of a larger and more complex organization, radically changing the nature of the role and level of responsibility, while typical employee duties may not be affected by the size of the company.

The Banks have increased in size significantly since 1995. In aggregate, the Banks have increased more than 315% in asset size and 739% in market capitalization between 1995 and 2012, relative to an increase in aggregate CEO compensation of 240% (in nominal terms) over the same time period.

Pay for Performance and Improved Governance

Higher pay levels have been associated with pay for performance because compensation is more at risk. The recent high level of focus on pay for performance has also increased CEO turnover, as boards are under increasing pressure and are more willing to fire underperforming CEOs.

[Another] explanation proposes that the growth in CEO pay is the result of stricter corporate governance and improved monitoring of CEOs by boards and large shareholders. Hermalin (2005) shows that, if CEO job stability is negatively affected by an increase in monitoring intensity, firms optimally respond by increasing CEO pay. According to this theory, the observed rise in pay should be accompanied by higher CEO turnover, a stronger link between CEO turnover and firm performance, and more external CEO hires. However, an economy-wide strengthening of governance may not lead to higher pay in equilibrium if the change also causes CEOs’ outside opportunities to become less appealing (Edmans & Gabaix 2010).

However, Taekjin Shin asserts that CEO pay (adjusted for performance) is higher at firms which create the appearance of shareholder value orientation. He suggests that firms adopt pay-for-performance monitoring and incentive-alignment mechanisms to create the appearance of shareholder value orientation, rather than to curb excessive executive compensation.

Increased CEO Mobility

Articles Suggesting CEO Mobility Increases Retention Risks

Carloa Frydman40 suggests that CEO mobility has increased as the importance of general management (rather than company specific) skills have increased. Her thesis is that:

Due to a decline in the monopsonistic power of firms and the better sorting of executives address firms, the model predicts a higher level of pay, higher within-firm inequality, higher variance in pay across corporations, and more mobility of managers across firms when the relative importance of general human capital increases.

Frydman attributes the shift in skills to:

  • Improvements in information and communication technology which have increased the scope, scale and complexity of firms
  • Improvements in modern finance and strategic analysis

She suggests that this increased executive mobility correlates with increased executive pay, increased inequality among top managers and increased turnover.

In support of her theory, Frydman notes that:

During the past three decades, the turnover of CEOs and other top management has significantly increased. In the 1970s about 15 percent of new CEO appointment in the firms in the Forbes annual surveys were outsiders, but 26.5 percent were hired from another firm in the 1990s (Murphy and Zabojnik 2004b). Moreover, the probability of forced CEO separation increased significantly over time, from 10.2 percent in the 1971 to 1976 period to 23.4 percent from 1989 to 1994 (Huson, Parrino and Starks 2001).

Frydman comments that:

The increase in external mobility during the past 30 years, coupled with the higher likelihood of moving later in the career, are consistent with an increase in the importance of general skills (and with the rising trends in executive pay).

In particular, Frydman notes that:

  • There has been an increase in the number of CEOs hired from outside and in the number of forced CEO transitions
  • There is in increase in late career transitions
  • There has been an increase in the years of formal education of CEOs
  • The number of CEOs with business degrees has increased significantly (with business degrees replacing technical and legal degrees)
  • The number of CEOs who have worked in different sectors of a firm (production, sales, human resources, finance and law) has increased

In her analysis, Frydman finds that:

…the level of executive pay, the inequality among top managers and the turnover of managers over their career are higher when general skills are more important.

Articles Suggesting CEO Retention Risk is Overstated

In contrast, as noted by NEI in its proposal, Elson and Ferrere assert that CEOs are less mobile and less frequently “poached” than commonly believed, resulting in pay which is higher than actually required for attraction and retention.

While generally acknowledging that the market is an appropriate basis for setting compensation for athletes, musicians and other superstars, the foundation of Elson and Ferrere’s argument is that the market should not determine executive pay as executives are not mobile and their talents are not transferable.

Elson and Ferrere, however, acknowledge that:

The concepts of supply and demand aid in an analytical determination of price in traditionally understood markets, but they are unable to provide precise conclusions where bargained outcomes may bear little relation to such competitive constraints and considerations. The market sets a floor and a ceiling on executive compensation, but, as we will argue, between which there is wide range for board discretion.

They also acknowledge that:

A chief executive, obviously, must earn more than their next best alternative employment opportunity would pay them, but they also earn less than the full value of their firm-specific productivity; the firm’s profit is similarly constrained between these payoffs. The precise determination of a wage (and firm profit) outcome must be determined within this flexible range through a complex bargaining process between the parties.

Canadian Banks’ Executive Mobility

The Banks have experienced no recent turnover among CEOs, as current CEOs have all been in place since at least 2007. Each CEO has spent more than a decade with the Bank, with most having spent multiple decades. However, the Banks have experienced meaningful movement among top executive ranks.

Not all of the Banks maintain records of the previous employer of new hires or the new employer of voluntary departures. Among those that were able to provide such records, there was clear evidence of the external competition for top talent faced by the Banks. Many senior level executives, including Executive Vice Presidents (senior officers likely included in the pipeline of potential CEO candidates), moved from one of the Banks to another. The Banks also hired executive talent from and lost executive talent to companies in the broader financial services sector.

The mobility of top executives among the Banks provides evidence of the need to include horizontal benchmarking assessments of executive pay levels. The Banks must understand the external market for talent that exists, as many of their top executives will have opportunities to join competitors. The Banks, therefore, should understand their executives’ “next best employment alternative” as described by Elson and Ferrere. There will also be circumstances where a Bank will need to pursue outside talent, and thus prepare a competitive compensation proposal sufficiently valuable to persuade an executive to leave their current role. While pay may not be the only or even the primary factor involved in turnover, it can play a very significant role.

Compensation Consultant Conflicts of Interest

Some critics have asserted that compensation consultants are partly responsible for the perceived excesses in executive pay. Higgins concluded in a 2007 study that companies using consultants offer significantly higher pay than companies not using consultants.

However, Murphy comments that:

…the cross-sectional correlation between CEO pay and the use of consultants does not imply that the consultants caused the high pay; it is equally plausible that companies with high pay are most likely to seek the advice of consultants. Indeed, Armstrong, et al. (2012) find no evidence of difference in pay between a sample of firms using consultants and a match sample of firms not using consultants. Similarly, based on a time series of 2006 to 2009 data, Murphy and Sandino (2012) find no evidence that firms increase pay after retaining consultants.

Concerns over the role of consultants led the SEC in 2006 to require disclosure aimed at identifying potential conflicts of interest of compensation consultants.56 Similarly, in Canada, public companies are required to disclose the fees paid to any compensation consultant providing advice to the Committee and the fees paid to such consultant for other work, similar to the requirements imposed on accounting firms providing audit services for the two most recent years.

What Alternatives Are There to Horizontal Benchmarking?

Various forms of vertical benchmarking have been discussed as potential alternatives to horizontal benchmarking. Vertical benchmarking involves evaluating the ratio of pay between top executives (typically the CEO) and other employees in the organization. Variations of vertical benchmarking include: the ratio between the CEO and all employees, the ratio between the CEO and other executives, and the ratio between the CEO and industry averages.

Vertical Benchmarking Ratios between Executives and All Employees

Vertical benchmarking generally evaluates the ratio of senior executive compensation relative to that of the broader workforce. This method gained particular attention in the U.S. with the passage of the Dodd-Frank Act in 2010.

Dodd-Frank Pay Ratio Disclosure

The Dodd-Frank Act compels the Securities and Exchange Commission (“SEC”) to implement rules requiring proxy disclosure of the ratio between the CEO’s total compensation and the median total compensation of the rest of that company’s workforce, with the total compensation calculation aligned with that used for the proxy circular summary compensation table. The SEC has yet to implement the provision, and many questions remain about how it will be implemented. The provision has been subject to significant debate, with much of the criticisms focused on whether the information will be useful to shareholders, how the information will be used by stakeholders and others and the difficulties and costs of calculating the median total compensation value.

Interestingly, the number of shareholder proposals respecting internal pay ratio disclosure among Russell

3000 companies in the U.S. dropped from 8 in 2010 to 3 in 2011, most likely as a result of the introduction of the Dodd-Frank Act. The proposals have generally received limited support; for example, the 2011 proposal by the International Brotherhood of DuPont Workers calling for disclosure of internal pay ratios received just 5.8% shareholder support.

Recently, the House Committee on Financial Services advanced a bill seeking to repeal this provision of the Dodd-Frank Act, although a repeal would likely face significant challenge gaining the approval of the Senate. If the provision is ultimately implemented, the disclosures will likely receive significant media scrutiny. The range of ratios between companies due to differences in workforces will make comparisons difficult, but companies will likely face political and media pressure related to changes in their ratio over time. It remains to be seen, however, to what extent shareholders will focus on the issue.

Concerns with the Dodd Frank Act Pay Ratio Disclosure

Many business groups have expressed concerns with disclosure of the pay ratio, as required under the Dodd Frank Act. A joint comment letter to the SEC from 23 business trade organizations asserts:

The corporate disclosure regime is designed to provide information that is useful to investors when making investment decisions. While it may be of general interest to some investors for much different purposes, it is unclear how the pay ratio disclosure will be material for the reasonable investor when making investment decisions. The ratio will inevitably vary widely among industries or businesses without any relevance to the financial performance of a company. Accordingly, some additional consideration of any possible benefit to be provided by this disclosure must be considered in the rulemaking process and weighed against the costs discussed below.

The submission discusses the costs associated with collecting employee compensation information for companies with multi-national operations. The concern about the cost of calculating the pay ratio is echoed by others and, given the complexity of the current provisions, which cross reference a number of other provisions of securities regulation, it is hard to entirely discount this argument.

David Hirschmann, President of the U.S. Chamber of Commerce’s Center for Capital Markets, echoed concerns with how the ratio would be used:

The ratio is not going to be a meaningful way to help investors but will be used as a political tool to attack companies.

The push by labor unions for disclosure of pay ratios supports the perspective that pay ratios will be used by stakeholders, to support their own agendas, such as the negotiation of pay increases in collective bargaining, which are not necessarily aligned with shareholder interests.

Expected Benefits of the Dodd Frank Act Pay Ratio Disclosure

Others, however, believe that the disclosure of the ratio will benefit shareholders. Robert Brown, Jr. made this case in the Harvard Business Law Review:

With respect to the materiality of the information, ratios have particular importance in the era of “Say on Pay.” An advisory vote on compensation gives shareholders an opportunity to comment on the reasonableness of CEO pay. Reasonableness, however, requires context. Metrics that allow for a comparison of pay practices among public companies can assist in providing the requisite context. Moreover, the ratios provide a mechanism for the first time for assessing the reasonableness of the compensation within each particular company.

Mr. Brown suggests that differences in business models, which impact pay ratios, can be explained in disclosure and that pay ratios will be a potential source of embarrassment to directors which will cause them to alter compensation to avoid adverse disclosure. See the discussion under the heading Government Policy Effects for a view on the effectiveness of these types of government intervention in the past.

Companies Using Vertical Benchmarking Ratios between Executives and All Employees

Few companies have established fixed limits on executive compensation as a ratio of overall workforce compensation, and changes in their approaches suggest the practice has made attracting and retaining executives challenging.

The historical compensation philosophy of Ben and Jerry’s Homemade, Inc. capped the salary and benefits (but not equity awards) of its executive officers based on a multiple of its lowest paid full-time worker with at least one year of service. In 1994, the last full year that founder Ben Cohen served as CEO, the limit was seven to one. However, the Board ended the policy when searching for a new CEO. The company’s proxy statement stated that the Board concluded that “the ratio had become a barrier to recruiting experienced people who can keep the Company true to, and successful at, realizing all three parts of its mission statement.” The Board instead implemented a policy of “linked prosperity” that included the use of more incentive pay to align management with shareholders.

For many years Herman Miller limited the CEO’s cash compensation to 20 times the average annual compensation earned by the Company’s regular full-time employees. However, the Company eliminated the cap in 1997, stating in its proxy statement that, in light of the amount of compensation at risk for its CEO, the Executive Compensation Committee believed it was inappropriate to maintain the cap. The Company’s former Board Chairman later stated,

From a competitive standpoint, we needed to eliminate the cap to attract and retain the right people.

Whole Foods continues to maintain a pay cap. In 1999, its policy limited cash compensation to 10 times the Company’s average annual wage. Whole Foods has had to increase the limit on multiple occasions. In 2000, the maximum multiple was increased to 14, where it remained until being increased to the current multiple of 19 in 2006. The CEO explained the later increase in a message to employees, stating that every top executive, except him, had been repeatedly approached by search firms seeking to lure them to rivals.

Each of these examples provides evidence of the challenge of attracting and retaining executive talent while using this vertical benchmarking approach as a primary determinant of executive pay. In each case, the cap did not include equity compensation, which often serves as the largest component of CEO compensation (more than 50% for each of the Banks). Additionally, each company found itself needing to eliminate or raise the limit on executive compensation to maintain the necessary executive talent to lead the organization.

Other smaller companies have disclosed information on the ratio of executive pay to that of the remaining workforce, but did not indicate any policy suggesting the ratio is used to limit pay. Northwestern Corp discloses the ratio of CEO pay (including salary and incentives) to the median pay of all full-time employees, and indicates that the ratio is monitored by the HR Committee. Similarly, El Paso Corporation (before being acquired in 2012) disclosed that its Compensation Committee monitored the relationship between total compensation for named executive officers and non-managerial employees. Southwestern Energy’s Compensation Committee reviews a 10-year history of the ratio between executive total compensation and that of lower-level employees when making its compensation decisions. Bank of South Carolina discloses the median salary for all employees other than the executive officers, but provides no commentary on whether the ratio is used by the Compensation Committee. MBIA previously disclosed average and median salary and bonus data for all employees other than named executive officers with no commentary on whether the information influenced decision making; however, it discontinued the practice in 2013.

Vertical Benchmarking Ratios between CEOs and Other Executives

Another form of vertical benchmarking assesses the ratio between the CEO’s compensation and that of other senior executives.

Some research suggests a relationship between higher ratios and lower corporate performance. One article correlates higher proportions of executive pay being delivered to the CEO with lower profitability and shareholder returns, as well as poor governance practices including reduced performance sensitivity of CEO turnover.

Moody’s includes an assessment of this internal pay equity ratio when incorporating executive pay into its credit analysis. The Company states:

A large ratio can be a possible sign of ‘key person risk’ or a weak board… In our view, a large disparity in internal pay equity (greater than three times the amount received by the executive second in pay, for example) may indicate underdevelopment of management succession planning, and concentration of power in the CEO. These factors pose a substantial succession planning challenge and some key person risk for the company at the CEO level. We think it is useful for boards to engage in this type of analysis as it provides a level of comfort that the board is exercising its role in succession planning and evaluation of the CEO’s progress.

Proxy advisors are also paying attention to this ratio. Institutional Shareholder Services lists an excessive ratio between CEO pay and that of the next-highest executive as a problematic pay practice. (Note that ISS did not express concern about this ratio at any of the Banks, as all had ratios under 1.5:1.)

A number of companies disclose that they consider the relationship between CEO pay and that of the other executives into their decision making. DuPont may be the most well-known example. The Company indicates that it manages the ratio between CEO pay and that of other Named Executive Officers, with the CEO’s total cash compensation targeted at a ratio between 2 and 3 times that of the other executives, and total direct compensation at a ratio between 3 and 4 times.

Other companies that disclose that this form of vertical benchmarking is an input into their decision making (but without disclosure of any fixed limits) include ConocoPhillips, Garmin, Intel, Kroger, PVH Corp, and Starbucks.

While this ratio may provide useful insight into the structure and governance of the top executive team, it does not address the primary concerns raised in the NEI proposal related to wage disparity between executives and the broader workforce.

Vertical Pay Ratios Relative to Industry Averages

Some external commentators have sought to present vertical pay ratios based on broader market average wage data. For example, Bloomberg recently published ratios of company specific CEO pay relative to U.S. government figures for industry-specific average pay and benefits.

While this form of vertical pay ratio avoids the need for each company to calculate it’s median or average pay level, several aspects of the approach limit its usefulness:

  • By focusing on broader market pay data, the ratio ignores the practices of the company regarding how its broader employee population is compensated.
  • Comparisons between companies ignore differences in company size and the corresponding differences in the scope of responsibility of executives. Generally, larger companies would be expected to pay higher CEO compensation which would result in a higher ratio.
  • By using a constant denominator for all companies in the same industry, in effect the comparison provides limited information beyond what the difference in CEO pay is between companies, which is already substantially disclosed in the summary compensation table of each company.

Horizontal Comparisons of Vertical Pay Ratios

Horizontal benchmarking can also be used by companies to review the appropriateness of their vertical pay ratios. In effect, market compensation surveys provide insight into the typical ratio between pay levels for different positions.

Lockheed Martin discloses that the Committee reviews the pay relationship between its NEOs, but uses horizontal benchmarking results to assess the appropriateness of the vertical pay ratios. Lockheed Martin’s disclosure states:

Consistent with its past practice…the Compensation Committee reviewed the pay relationship of the NEOs…Because the principal elements of our compensation program are based on the market rate, our internal pay equity reflects the relative pay of our comparator group of companies.

Conclusion

Vertical benchmarking can provide additional context for Committees establishing executive compensation levels, but it is not without its limitations.

Companies that have attempted to establish fixed executive pay limits based on the ratio of CEO pay relative to the broader workforce have found the application of those limits challenging. The review of such a ratio, however, may provide useful context for Committees, particularly in assessing trends in pay disparity. The following section includes recommended guidelines for companies to develop a useful ratio.

Other vertical benchmarking methods are unlikely to be useful in addressing pay disparity. Benchmarking the ratio of CEO pay to that of other executives may provide insight into succession planning and the dynamics of the executive team, but will not reflect compensation of the broader workforce. Ratios based on broader industry averages ignore company-specific pay practices for employees other than the CEO, and will only reflect variations in CEO pay between companies.

What Safeguards and Practices Should Be Followed when Using Horizontal Benchmarking for Executive Pay?

While vertical benchmarking can provide important additional context to assist Compensation Committees in their decision making, it remains critical for companies to review horizontal benchmarking data to understand the competitive market place for talent. Implementing appropriate safeguards within the horizontal benchmarking process ensures that data are considered in proper context to avoid inappropriate escalations in executive pay levels.

Broadly speaking, safeguards can be viewed in two categories: (i) those related specifically to the development and use of horizontal benchmarking data, and (ii) those related to broader issues of strong corporate governance to oversee the decision making process.

Safeguards on the Development and Use of Horizontal Benchmarking Data

Selection of Appropriate Peer Groups

Committees should carefully evaluate the selection of peer companies. The peer companies should be appropriate in terms of size, nature of the business and geography, to prevent compensation from being compared against the compensation of companies that are significantly larger or pay on a different basis (e.g., a retail bank would not generally compare its compensation to an investment bank). If a lack of available relevant competitors leads to broader and less ideal selection criteria, the Committee should understand the group’s limitations and how they impact the resulting compensation data.

Canadian companies frequently consider whether it is appropriate to include U.S. companies in a compensation peer group. The inclusion of U.S. companies presents challenges due to the structural differences in compensation, including incentive plan design due to differences in tax law and higher leverage (upside and downside) in U.S. incentive arrangements, and because the pay levels at U.S. companies tend to be higher than at their Canadian counterparts. Generally companies will include U.S. companies in their peer group if they have significant U.S. operations or sales, U.S. based direct competitors for talent, and/or few comparable Canadian companies. While in many circumstances it may be appropriate to include U.S. peers, Committees should monitor the impact U.S. companies have on benchmarking data and compensation levels.

Ensure Integrity of Market Data

Committees should ensure they can have confidence in the market data they are using to make decisions. Several factors are important:

  • Any premiums or discounts applied to the market data based on an incumbent’s responsibilities relative to survey benchmarks should be critically evaluated.
  • The methodology for valuing long-term incentive grants among peer companies should be consistent with how the company determines the value of its awards.
  • Market data should typically exclude any special one-time awards, such as new hire grants (including buyouts of previous employer awards) and retention grants.
  • Executive pay should be compared to market in total and not just on an individual component basis. The sum of market pay levels for individual components (i.e., base salaries, short-term incentives, and long-term incentives) will not necessarily match market pay levels for total pay due to differences in pay mix among peers.

Avoid Targeting Above Median Compensation

Absent a compelling rationale for a different approach, Committees should seek to provide target compensation opportunities within a competitive range of the market median. While there are occasions when it may be appropriate to target above median pay, such as when a majority of peers are significantly smaller, Committees should be very cautious before approving a philosophy of targeting above median pay. While target pay opportunities should generally be aligned with market median pay opportunities, executive pay programs should be structured to allow actual pay to vary above or below median levels based on executive and company performance.

Evaluate Consistency of Executive Pay Actions with Pay Decisions for the Broader Company

Committees should evaluate the consistency of executive pay actions with those proposed by management for the rest of the company. For example, salary increases provided to executives should be compared to the merit budget for the rest of the organization. Likewise, Committees can assess the payout from executive bonus plans to those for other employees. These evaluations can provide Committees with insight into how the Company’s resources are being shared at different levels of the organization.

Evaluating trends in vertical benchmarking ratios can also be a useful tool to assess the consistency of compensation decisions throughout an organization. The Banks can develop calculations of vertical benchmarking that provide meaningful information while avoiding the complexity of the pay ratio disclosure provisions within the Dodd-Frank Act. While there are no established best practices for vertical benchmarking given its limited use, following are suggested guidelines for developing a useful ratio:

  • Focus on developing an approach to calculating a ratio that can be consistent over time, avoids complexity, and uses data that is readily attainable.
  • Define a population of the broader employee workforce included in the ratio calculation. For example, a company could choose to include all non-executive full-time employees. Alternatively, the population could include a sample of key jobs throughout the organization. The population can avoid groups that pose data collection and calculation challenges, such as international employees and which can be used consistently year over year.
  • Define the executive population included. This could be just the CEO, or could include the broader executive team.
  • Establish a definition of compensation that provides a full picture, can be easily calculated and can be used consistently year over year.
    • For example, compensation could include salaries, annual bonuses, and long-term incentives. Given their primacy in executive compensation, long-term incentives should be included in the ratio to provide a complete picture of compensation.
    • Including incentives based on target opportunities (as opposed to actual payouts) would avoid ratio fluxuations based on performance due to the increased leverage in executive pay arrangements; however, many organizations do not have established target bonus opportunities for all incentive- eligible roles.
    • Calculating annualized values for company-provided benefits for all of the included employees may be overly complex. Benefits are generally a smaller percentage of pay for executives than for the broader employee population. However, Committees may want to consider how benefits may impact pay disparity.
  • Avoid prescriptive limits on the ratio, but evaluate and understand trends over time. Companies that have attempted to set vertical pay ratio caps have struggled to maintain consistent limits. Company circumstances can change over time, including the size of the organization and the makeup of the employee population. However, Committees can review movement in the ratio and evaluate whether or not changes in the ratio are appropriate.

Avoid Literal Application of Market Data

While benchmarking is an important input into compensation-setting decisions, Committees should avoid the temptation to apply it literally, setting pay precisely in line with the market data without applying business judgment. Pay decisions should be made considering a number of factors, including:

  • Assessment of real retention risk
  • Consideration of the scope of the role
  • Seniority and experience of the incumbent
  • Long-term performance of the incumbent
  • Company’s business and strategy

Committees should ensure that all of this information is provided to them in the context of executive pay setting.

Ensuring Strong Corporate Governance of Executive Pay Decisions

Independent and Qualified Board Oversight

There are a number of good governance practices that companies can follow in establishing oversight of executive pay, including:

  • Ensuring that the Compensation Committee is comprised solely of independent directors
  • Ensuring that Committee members are fully compensation literate
  • Ensuring independent Board leadership by separating the roles of chair and CEO or establishing a strong lead director with a significant mandate
  • Active engagement by the Committee in considering executive compensation and setting pay levels

Succession Planning

Effective internal development and succession planning can help organizations to manage compensation of executives. High CEO pay has been attributed in part to the increased prevalence of hiring outside CEOs, exacerbated by hiring practices, including having the new CEO’s contract negotiated and drafted by the internal counsel.

External hires, while they may be appropriate at times, carry several risks for the organization. Elson notes that research suggests a negative expected benefit on performance from external CEO hires. There are also compensation related risks. External candidates typically have higher leverage in negotiations which can lead to higher pay and may require an organization to go outside its desired compensation structure. External candidates also must typically be compensated for outstanding incentive awards and retirement benefits that will be forfeited if they leave their current organization.

On the other hand, internal promotions typically receive initial compensation below market median levels, reflecting their inexperience in the role. Internal candidates will also likely not require changes to a company’s preferred compensation structure, sign on bonuses or make whole payments.

Evaluate the Pay and Performance Relationship

Shareholders are generally more concerned with the alignment of pay and performance than with pay levels. Accordingly, in addition to reviewing pay levels, compensation committees should carefully consider how their incentive plans work and should:

  • Consider choice of metrics, to ensure that metrics are aligned with and support the company’s strategic and budget objectives
  • Review the level of stretch in performance targets, by reference to internal and external and relative and absolute standards
  • Stress test and back test performance and metrics
  • Align executive compensation decisions with the decisions made for the broader employee community (i.e., having broad-based incentive (particularly annual incentive) plans with companywide participation)

Independent Compensation Advice to the Committee

The Canadian Coalition for Good Governance, in its 2013 Executive Compensation Principles, suggests that having an independent consultant provide executive compensation advice to the Committee is best practice. As discussed previously, while some critics have suggested that compensation consultants may contribute to increasing executive pay, there is no clear evidence supporting the claim. That said, there is the potential for a conflict of interest when executive compensation advice is provided by a full-service consulting firm, the bulk of whose work with a client involves pension and other services for which they are retained by management. This potential conflict has been recognized and the practice of having independent executive compensation advice provided to the Committee has very significantly increased.

The Dodd-Frank Act in the U.S. expanded the requirements related to consultant independence. The requirements do not establish a litmus test for independence; rather, Compensation Committees must make their own determination based on a consideration of six prescribed factors.68 Companies must disclose that the Committee considered the factors, whether or not they uncovered any conflicts, and if so, how those conflicts are being addressed.

Canadian disclosure requirements focus on the type and cost of services the Committee’s consultant provides to the Committee and management.

Communicate with Shareholders through Say on Pay and Related Dialogues

Although Say on Pay votes do not bind compensation committees to make program changes, they do present shareholders with an opportunity to express concerns respecting executive compensation and provide a means of holding Compensation Committees accountable for their decisions. The percentage of votes received in favour is indicative of overall shareholder sentiment about company compensation programs.

Meridian compared the percentage of votes in favour at each of the six largest Canadian banks with the average percentage of votes in favor both in Canada and in the U.S. for 2011 voting, 2012 voting and 2013 voting (up to the release of this report). The chart below depicts the percentage of votes received in favour of Say on Pay for each of the six largest Canadian banks in 2011, 2012 and 2013, along with the average percentage of votes received in Canada and the U.S. for each of those years.

Historical Results of Canadian Banks’ Say on Pay Votes
Bank Say on Pay Vote Results (% in Favour)
2013 2012 2011
Bank of Montreal 96.90% 90.90% 92.80%
Bank of Nova Scotia 94.40% 92.60% 96.80%
Canadian Imperial Bank of Commerce 96.60% 97.50% 97.30%
National Bank of Canada 96.60% 95.50% 97.50%
Royal Bank of Canada 93.80% 84.90% 88.60%
Toronto-Dominion Bank 94.50% 93.70% 96.60%
Canadian Average 89.20% 91.80% 93.90%
U.S. Average 91.30% 90.20% 90.90%

The high say on pay results received by the Banks suggests that shareholders have generally been in favour of the Banks’ executive pay practices. Per the statistics compiled thus far for 2013, each of the Banks has received significantly more support than both the Canadian average and the U.S. average.

Concluding Comments

Horizontal benchmarking will continue to be an important tool for Committees as they make decisions regarding executive compensation. While there is no compelling argument that horizontal benchmarking has been a primary contributor to excessive CEO pay, it can lead to excessive pay increases when used inappropriately. A number of safeguards can help ensure a proper application of horizontal benchmarking, including evaluating how pay decisions for executives compare to those made for the broader workforce. While vertical benchmarking has its limitations, an appropriately designed ratio of CEO pay to other employees can provide useful context for Committees in their evaluation. With both horizontal and vertical benchmarking, it is important for Committees to avoid overly prescriptive applications of data and instead apply business judgment to ensure compensation actions are appropriate for the given circumstances.