Corporate Finance VIII: Corporate Governance

Slide 2

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 3

OUTLINE

FACTS
- U.S. CEOs
- The rest of the world
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 4

FACTS: U.S. CEOS - HOW: SALARY & PERFORMANCE BONUSES FOR U.S. CEOS

Salary: (<1 Million)
Bonus: usually an 80/120 plan with modest accounting-based targets

Graph of a typical CEO bonus plan structure.

Description of the Graph:

This graph illustrates a typical performance bonus structure for a CEO, often called an "80/120 plan". The x-axis represents an accounting metric (labeled "Acct. #"), such as earnings per share. The y-axis represents the "Bonus Payment".

The graph shows a red line with three distinct segments:

A flat horizontal line at a low level (often zero) for performance below a certain threshold, marked as "80% x T*", where T* is the target performance level.
A straight, upward-sloping line between 80% of the target and 120% of the target. In this range, the bonus payment increases linearly with performance.
Another flat horizontal line at a maximum level for performance above the cap, marked as "120% x T*". This indicates the bonus is capped once performance significantly exceeds the target.

Slide 5

HOW: STOCK OPTIONS

Stock Options: usually at-the-money with 10 year expiration

Description of the Graph:

This is a classic "hockey stick" diagram illustrating the payoff of a stock option at expiration. The x-axis represents the "Stock price" and the y-axis represents the "Option Payoff".

A red line shows the payoff profile:

The line is flat at zero for all stock prices below the exercise price (labeled "Ex. price in 10 yrs." and also indicated by "Current price" for an at-the-money option). This region is labeled "out-of-the-money".
At the exercise price, the line kinks upwards and then rises linearly with a slope of 1. For every dollar the stock price is above the exercise price, the option payoff increases by one dollar. This region is labeled "in-the-money".

This shows that the option holder only profits if the stock price rises above the price at which the option was granted.

Slide 6

HOW: RESTRICTED STOCK AND SEVERANCE PACKAGES

Restricted Stock: Equivalent to stock option with zero exercise price
Long-term incentive pay (structure varies with firm; frequently complex and industry specific)
Pension plans/severance packages: SERPs

Slide 7

HOW: RELATIVE PERFORMANCE EVALUATION (RPE) PLANS

A peer comparison group for the CEO is defined
Rewards from the RPE plan are based on the CEO's percentile ranking within her peer comparison group with respect to
- Earnings
- Total shareholder returns (TSR)
- Other: Revenue, costs...
Use of RPE compensation is common and has (possibly) expanded greatly in last 10-15 years.

Slide 8

OUTLINE

FACTS
- U.S. CEOs
- The rest of the world
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 9

FACTS: THE REST OF THE WORLD - HOW ELSEWHERE

Stock based compensation less important outside the U.S: Approximately 50% lower in the UK; though option-based compensation was established earlier in the UK.
In UK performance shares, restricted stock grants conditional on meeting performance targets become dominant stock-based form of compensation.
Stock-based compensation even less prevalent in the Eurozone, and Asia.

Slide 10

CEO COMPENSATION AROUND THE WORLD

Stacked bar chart of CEO compensation across various countries.

Description of the Graph:

This is a stacked bar chart showing average CEO compensation across numerous countries. The y-axis represents total compensation in USD, ranging from 0 to 3,000,000. Each bar represents a different country.

The bars are composed of four stacked segments, representing different types of pay (from bottom to top):

Salary (dark blue)
Other pay (light blue)
Non-equity incentive pay (orange)
Equity incentive pay (yellow)

The most striking feature is that the bar for the United States (far left) is significantly taller than all others, exceeding $2,500,000. The composition is also different, with a very large portion coming from equity and non-equity incentive pay. Other countries like Switzerland, Canada, and the Netherlands follow, but at much lower total levels. Countries in Asia, such as China, Malaysia, and India, are shown at the far right with the lowest levels of total compensation.

FIGURE: CEO compensation around the world: From "The Executive Compensation Controversy: A Transatlantic Analysis," by Martin J. Conyon, Nuno Fernandes, Miguel A. Ferreira, Pedro Matos, and Kevin Murphy.

Slide 11

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 12

TRADITIONAL THEORIES

These frameworks focus on the shareholder/manager relationship within firms
The board acts as agent representing shareholder interests in negotiations with management
Principle/agent model considers how compensation will be determined if the board acts in the interests of shareholders
The skimming model consider how compensation will be determined if the board is suborned by the CEO

Slide 13

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
- Principal/Agent (P/A) theory
- Theories of Compensation: Skimming
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 14

IDEA BEHIND PRINCIPAL/AGENT (P/A) MODEL

Managers actions affect firm value
Managers actions frequently cannot be observed or verified by shareholders
Since actions cannot be observed contracts must be written on output
However, output is an imperfect proxy for actions as output is affected by luck
This imperfection leads to agency-costs to shareholders:
- reduced efficiency of operations
- increased managerial compensation

Slide 15

A SIMPLE P/A MODEL

Manager hired to manage an asset
Manager's effort is not verifiable but the firm's cashflow is observable and verifiable
The effort of the manager affects the cashflow produced by the asset
The more effort the manager makes, the greater the likelihood of a high cashflow
But, even with high effort, a high cashflow is not assured

Slide 16

ASSUMPTIONS

Firm lasts for one period, at the end of the period a cashflow is received - High or Low
Manager will receive a compensation package that is a function of the cashflow
Compensation package satisfies limited liability-the manager's compensation is never negative and is always less than the whole cash flow.
Manager and shareholders are risk-neutral.
Risk-free rate of interest is 0.
The manager does not have to work for the firm and can earn a cash payment of w from alternative employment
The manager experiences a nonpecuniary cost of c from providing High effort and 0 from providing Low effort
The manager's effort is not observable

Slide 17

EFFORT AND CASH FLOWS

Relation between managerial effort and firm cash flow
Manager's Effort	$\mathbb{P}[CF=v_H]$	$\mathbb{P}[CF=v_L]$
High	$p_H$	$1-p_H$
Low	$p_L$	$1-p_L$

Slide 18

MINIMAL COMPENSATION TO ENSURE HIGH EFFORT

Any payment to the manager which is the same for both High and Low cash flows, will not encourage effort
Because, the manager is risk neutral, he places the same value on all compensation packages that produce the same expected payment.
Thus, the least expensive compensation package that ensures effort will not have a fixed component. Let i represent the variable payment to the manager, made if and only if the cash flow is High.
In order for the manager to exert High effort, the manager's payoff from High effort must at least equal the managers payoff from low effort

$$ \underbrace{p_H i + (1-p_H)0 - c}_{\text{Pay-H}} \ge \underbrace{p_L i + (1-p_L)0}_{\text{Pay-L}} $$

Slide 19

EFFORT ASSURING COMPENSATION

Or,

$$ i \ge \frac{c}{p_H - p_L} $$

The firm will chose the smallest value of i that satisfies this condition, so the optimal compensation for ensuring high effort is

$$ i^* = \frac{c}{p_H - p_L} \quad (1) $$

Slide 20

VALUE OF HIGH-EFFORT ASSURING COMPENSATION

If the manager is paid enough to secure effort, the expected value of managerial compensation is $i^* p_H$. Using Equation (1) we see that the value of the smallest compensation package that secures effort is $p_H \frac{c}{p_H - p_L}$
Effort cost the manager is c.
Thus, the total payoff from accepting this package equals $p_H \frac{c}{p_H - p_L} - c$

Slide 21

COMPENSATION FOR HIGH EFFORT

Thus, the manager will accept the High-effort assuring package only his payoff at least equals his payoff from alternative employment:

$$ p_H \frac{c}{p_H - p_L} \ge w+c $$

If this condition is not satisfied, the incentive payment will have to be increased until the value of the manager's contract reaches his reservation wage plus the cost of effort.
Thus, if High effort is assured, expected managerial compensation equals

$$ \max\left[p_H \frac{c}{p_H - p_L}, w+c\right] \quad (2) $$

Slide 22

COMPENSATION FOR LOW EFFORT

The only requirement for compensation that ensures Low effort is that its total value of compensation equals the manager's reservation wage.
Thus the cost of assuring low effort is simply w.

Slide 23

OPTIMAL POLICY: HIGH OR LOW EFFORT?

If the shareholders assure Low effort, their payoff is $ShPayLow = p_L v_H + (1-p_L)v_L - w$
If shareholders assure High effort, their payoff is $ShPayHigh = p_H v_H + (1-p_H)v_L - \max[p_H \frac{c}{p_H-p_L}, w+c]$
Shareholders will choose compensation sufficient to ensure High effort whenever ShPayHigh > ShPayLow. If the reverse inequality holds, shareholders will accept low effort.

Slide 24

OPTIMAL POLICY WITH NO P/A PROBLEM

In the absence of a agency problem, the manager's effort is observable and verifiable. The firm must pay the manager $w+c$ if it contracts him to exert high effort; If the firm is willing to accept low effort then it need only pay the manager w. Thus the firm will prefer to induce effort whenever

$$ (v_H - v_L)(p_H - p_L) > c $$

the reverse inequality obtains the firm will prefer to induce low effort even when there is no P/A problem

The optimal employment contract in the absence of a P/A problem, is called the first-best solution.

Slide 25

DISTORTIONS FROM AGENCY PROBLEM

As you can see from (2), the presence of an agency problem increases the cost of ensuring High effort (relative to the first-best solution) by

$$ \max[p_H \frac{c}{p_H-p_L} - (w+c), 0] $$

This increased cost leads to two potential distortions:
- Inefficiency: Because of the increase in compensation required to induce High effort, firm does not provide compensation that induces High effort, even though High effort is first best.
- Managerial Rents: The firm opts for High effort when High effort is first best, but is forced to pay the manager more than reservation demand for High effort, $w+c$

Slide 26

EFFECTS OF AGENCY CONFLICTS

The effect of agency conflicts depends on the firm's losses from reduced effort and amount of noise in the effort/performance relation

OP: The manager's effort is the same as it would have been absent agency conflicts, but his compensation is inflated because of agency conflicts.
LE: The level of effort is lowered by the agency conflict. The manager's payoff is not effected by the agency conflict.
H: The agency problem is not severe enough to affect compensation or effort. Both the first best and agency solutions call for High effort.
L: The agency problem is not severe enough to affect compensation or effort. Both the first best and agency solutions call for Low effort.

Slide 27

AGENCY COSTS

Graph showing different agency cost outcomes based on model parameters.

Description of the Graph:

This graph shows how different agency problem outcomes arise depending on the underlying parameters of the firm. The x-axis is the ratio $p_L/p_H$, representing how informative performance is about effort (a lower ratio means performance is a better signal of high effort). The y-axis is the ratio $v_L/v_H$, representing the downside risk of the project (a lower ratio means a bigger drop in cash flow in the bad state).

The parameter space is divided into four colored regions, each corresponding to a different outcome:

H (green, bottom left): Agency problems are minimal. It's optimal to induce High effort, and it can be done without paying extra rents.
OP (blue, bottom center): "Overpayment". It's still optimal to induce High effort, but because performance is a noisy signal, the firm must pay the manager "rents" (compensation above their reservation wage) to incentivize them.
LE (yellow, middle): "Low Effort". The agency problem is so severe (performance is a poor signal of effort) that it becomes too expensive to incentivize high effort. The firm settles for low effort, even though high effort would have been optimal without the agency problem.
L (red, top right): Low effort is optimal even in a first-best world, so the agency problem doesn't change the outcome.

FIGURE: Parameters for the example are as follows: $v_H=3$, $p_H=0.75$, $c=0.20$, and $w=0.05$. The other parameters, $v_L$ and $p_L$ are varied in the graph.

Slide 28

OBSERVATIONS

In the P/A model firms will "overpay" managers in some cases even when shareholders control the compensation decision and managers have no influence on boards.
Changes in the model parameters that are outside the managers' control can increase managerial compensation: there is "pay for luck"
The agency problem is costly for shareholders but can benefit managers.
Managers earn the highest rents (excess compensation) when the relation between managerial effort and output is moderate, not too strong and not to weak.

Slide 29

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
- Principal/Agent (P/A) theory
- Theories of Compensation: Skimming
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 30

"AGENTS WITH OR WITHOUT PRINCIPALS" (BERTRAND & MULLAINATHAN, 2000)

In the P/A, shareholders set CEO benefits and salaries to maximize firm value
An alternative perspective: skimming-"CEOs set their own pay."
CEOs use their influence over corporate boards, particularly the compensation committee to ensure that they are paid more than they are worth
Motivation for theory
- In mid-1980s, US passed laws to make hostile takeovers more difficult, reducing the risk of takeover in poor performance
Entrenched CEOs, no longer constrained by the threat of hostile takeovers, were free to use their influence over boards for self enrichment

Slide 31

NUMBER OF HOSTILE BIDS

Line graph showing the number of hostile takeover bids over time for several countries.

Description of the Graph:

This is a time-series line graph showing the number of hostile takeover bids from 1978 to 2010. The x-axis is the year, and the y-axis is the number of bids, from 0 to 100.

There are lines for five countries. The most prominent is the blue line for the United States, which shows a dramatic surge in hostile bids starting around 1980, peaking in the late 1980s (at nearly 90 bids), and then sharply declining in the early 1990s. The green line for the United Kingdom shows a similar, though less pronounced, pattern. The lines for Germany, France, and Japan remain near zero for the entire period. This graph illustrates the "takeover wave" of the 1980s in the US and UK and its subsequent decline, which is a key institutional change discussed in the context of the "skimming" theory of CEO pay.

Source: SDC Platinum.

Slide 32

"AGENTS WITH OR WITHOUT PRINCIPALS" - II

B&M look at firms with different corporate governance structures in the US.
Blockholders: shareholders with > 5% of shares.
They observe that reforms induce:
- Large increase in mean pay for firms without block shareholders.
- Large increase in performance-based pay for firms with block shareholders.
They argue that if the CEO is 'lucky', e.g., when demand conditions are favourable:
- Agency theory tells us that pay should not rise.
- 'Skimming' view tells us that it should.
They find: CEO pay is sensitive to luck. But block shareholders block pay rises related to luck.

Slide 33

PROBLEMS WITH SKIMMING THEORY: I

If CEO pay is determined by CEO influence over the board
We should expect that entrenched CEOs earn more than newly hired CEOs and entrenchment would prevent the forced removal of underperforming CEOs
- But big packages are negatively associated with CEO tenure (Murphy, 2002)
- Average CEO tenure fell from 10 years to 4 years over the period (80s-00s) where CEOs supposedly started manipulating boards (Henderson, 2007)

Slide 34

PROBLEMS WITH SKIMMING THEORY: II

CEOs should also be able to use their entrenchment to increase their compensation relative to other elite workers (e.g., law firm partners, private equity and hedge fund managers) who cannot "set their own pay"
- But this has not happened:
- Relative to ordinary workers CEO compensation has greatly increased but CEO compensation has not increased significantly relative to other elite workers (Kaplan, 2008)
Pay for luck is not inconsistent with the P/A model so finding a pay for luck relationship cannot per se identify skimming

Slide 35

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 36

NEWER THEORIES

Over the last few decades, general management human capital has become a more important component of managerial human capital
Increased general managerial human capital increases the substitutability of managers within and across firms
Leading to more appointments of external candidates, more turnover, and more active managerial labour markets
The newer theories focus on the effect of labour markets on CEO compensation packages
Tournament theory focuses on the incentive effects
Positive assortative matching (PAM) models focus on how the outside-the-firm employment options of managers affect compensation

Slide 37

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
- Tournaments
- Positive Assortative Matching (PAM)
GOVERNANCE
SUMMARY

Slide 38

TOURNAMENT THEORY: IDEA

Principals frequently lack the information required to set reasonable output based contracts (e.g., $i^*$ in the P/A model above)
Even when such information is available, principals have ex post incentives to lower compensation by underestimating absolute levels of agent performance.
Tournament theory solution: let agents compete with each other, rewarding the best performer(s) through tournaments.

Slide 39

COSTS OF TOURNAMENT COMPETITION: RISK-TAKING

When competing for rank-based rewards
Competitors have an incentive to adopt risky "lose big/win small" strategies
- If your performance is better than a rivals, you will outrank a rival regardless of whether you just eke out a slightly better performance, or far surpass the rival
- If your performance is worse than a rival's, you will be outranked regardless of whether you your performance is a bit worse or a lot worse
"lose big/win small" strategies entail ruin risk-taking
- accepting a small probability of a huge loss in exchange for higher probability of better than average performance
- For example, a hedge fund writing out-of-the-money puts on the S&P 500

Slide 40

COSTS OF TOURNAMENT COMPETITION: SABOTAGE

Sabotage: Investing effort in lowering rivals reported or actual performance rather than increasing own performance,
- e.g., office politics, withholding information, negative disclosures about rival CEOs' firms
Evidence
- CEO Steve Ballmer's introduction of forced ranking performance evaluation may have lead managers to withhold information from colleagues and adversely affected Microsoft's performance (Oremus, Slate, 2013)
- German professional football: Underdog teams are much more likely to win tackles by fouling than superior teams (Deutscher, et al., 2013)
- Balafoutas and Sutter (2012) document similar results in Judo tournaments.

Slide 41

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
- Tournaments
- Positive Assortative Matching (PAM)
GOVERNANCE
SUMMARY

Slide 42

POSITIVE ASSORTATIVE MATCHING (PAM)

Individual CEOs are part of a managerial labour market
CEOs are substitutes, so the compensation of each CEO affects the compensation of others
Firm quality and CEO quality are strategic complements: The marginal value of increased CEO quality is higher for higher quality firms.
This leads to positive assortative matching: The best CEOs are hired by the biggest and best firms.

Slide 43

MATCHING: EXAMPLE

Suppose we have a market for CEOs with two job candidates: Miss Great and Mr. Good
There are two firms who might hire these candidate CEOs: Great Firm and Good Firm
Firm value under the two CEOs (gross of CEO compensation) are given by the following table:

Firm value gross of CEO compensation
	Firms
CEOs	Good	Great
Good	30.000	34.500
Great	37.50	43.125

Slide 44

MATCHING EQUILIBRIUM

A positive assortative matching (PAM) is an assignment of the best CEOs to the best firms
PAM equilibrium is a salary for Mr. Good and Miss Great that satisfies the following conditions:
- No firm can offer another firm's CEO compensation that
  - will be accepted by the other firm's CEO and
  - result in an increase in the offering firm's value

Slide 45

EQUILIBRIUM EQUATIONS: EXAMPLE

The salaries of the Great and Good managers, $s_{Great}$ and $s_{Good}$ in a positive assortative matching must satisfy the following equations:
The Great firm cannot gain from hiring away the good manager at a salary higher than the Good manager is already receiving $ \underbrace{43.125 - s_{Great}}_{\text{GreatMgr/GreatFirm}} \ge \underbrace{34.500 - s_{Good}}_{\text{GoodMgr/GreatFirm}} $
The Good firm cannot gain from hiring away the Great manager at a salary higher than the Great manager is already receiving $ \underbrace{30.00 - s_{Good}}_{\text{GoodMgr/GoodFirm}} \ge \underbrace{37.50 - s_{Great}}_{\text{GreatMgr/GoodFirm}} $

Slide 46

GRAPH OF EQUILIBRIUM COMPENSATION

Graph showing the set of equilibrium CEO salaries.

Description of the Graph:

This graph shows the set of possible equilibrium salaries in the two-manager, two-firm example. The x-axis is the salary of the "Good" CEO, $s_{Good}$, and the y-axis is the salary of the "Great" CEO, $s_{Great}$. Both axes range from 0 to 40.

A single, thick, colored line is plotted, running diagonally from the bottom-left to the top-right. This line represents the set of salary pairs ($s_{Good}$, $s_{Great}$) that can exist in a stable, positive assortative matching equilibrium. Any point on this line is a possible outcome where neither firm has an incentive to poach the other's CEO. The line shows that the salaries of the two CEOs are positively correlated; if one gets paid more, the other must also be paid more to maintain the equilibrium.

FIGURE: CEO equilibrium compensation for the two CEOs

Slide 47

GENERALIZING THESE IDEAS

Consider a world with ten firms and ten managers
The ability of manager i, $a_i$, equals i
The asset base of firm k, $b_k$, equals k
The output of firm k when managed by manager i equals $b_k a_i$

Slide 48

POSITIVE ASSORTATIVE MATCHING

Suppose our aim is to maximize total output. How should we match up managers and firms?
Total output if we match up the biggest firms with the best managers: $b_1 a_1 + b_2 a_2 + b_3 a_3 + \dots + b_9 a_9 + b_{10} a_{10} = 1 + 2^2 + 3^2 + \dots + 9^2 + 10^2 = 385$
Total output if we match up the biggest firms with the worst managers: $b_{10} a_1 + b_9 a_2 + b_8 a_3 + \dots + b_2 a_9 + b_1 a_{10} = 10\times1 + 9\times2 + 8\times3 + \dots + 2\times9 + 1\times10 = 220$
Output is much higher when the matching is positive assortative, i.e. the biggest firms are matched with the best managers

Slide 49

PAM EQUILIBRIUM

In general, the Pareto efficient matching is the positive assortative matching
The basic PAM model is neo-classical, no market imperfections
So, as you learned in your Micro class, by the First Welfare Theorem, in a PAM equilibrium in the managerial labour market, matching is positive assortative

Slide 50

WHAT DOES THIS HAVE TO DO WITH CEO COMPENSATION?

Consider manager 2: manager 2 ability is 1 unit greater than manager 1's
- How much higher is firm 2's output if firm 2 hires manager 2 rather than manager 1? $\underbrace{2\times2}_{\text{hire mgr.2}} - \underbrace{2\times1}_{\text{hire mgr.1}} = 2$
Now consider manager 10: manager 10 ability is 1 unit greater than manager 9's
- How much higher is firm 10's output if firm 10 hires manager 10 rather than manager 9? $\underbrace{10\times10}_{\text{hire mgr.10}} - \underbrace{10\times9}_{\text{hire mgr.9}} = 10$

Slide 51

MORAL OF THE STORY

The gap between manager 2 and manager 1's ability equals the gap between manager 10 and manager 9's ability.
However the value this ability difference is 5x greater for firm 10 (manager 10's employer) than firm 2 (manager 2's employer)
Thus, the gap between the compensation of managers 10 and 9 should be much larger than the gap between managers 2 and 1, even though the ability differences are the same
Therefore PAM models predict that CEOs of S&P500 and FTSE firms are in the very top range managerial ability and thus they can command huge compensation premia for slightly greater ability

Slide 52

IMPLICATIONS OF PAM

Sudden increases in the values (b increases) of the firms lead to higher compensation even if CEO talent is fixed: i.e., we have "pay for luck" even though CEOs do not suborn the board.
Benchmarking the pay of one CEO to the other is perfectly rational and in the interests of shareholders.
The distribution of CEO compensation is positively skewed, i.e. mean compensation is significantly greater than median compensation
Pay based on relative performance RPE makes sense for the firm: managers' values equal the amount rival firms will pay to poach them

Slide 53

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 54

GOVERNANCE

How do owners ensure that firms are operated in their interests? Governance
Governance is shaped by formal institutions, laws, accounting conventions, exchange listing rules, etc.
But how these institutions are operated in practice varies with the structure of share ownership, cultural norms, the efficiency of capital markets
- Example: After WWII, Japanese bankruptcy law institutions were largely borrowed from US bankruptcy law
- But the typical way financial distress is resolved in Japan is very different from distress resolution in the U.S.

Slide 55

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
- Formal institutions
- How governance actually operates
SUMMARY

Slide 56

CORPORATE BOARDS: WHAT?

Legal Role: monitor management decisions, ratify the most important, evaluate, hire, and replace top management, advise top management, insure compliance with laws and regulations
Legal duties of board members: Loyalty, Care
Rights and protection: Business judgement rule

Slide 57

FEATURES

Interlocking is common
- the boards of most large (> $ Billion) companies include at least one CEO of another company
- A 1/5 of directors serve on four or more boards.
28% of firms have at least one director who owns no shares of the company's stock.

Slide 58

FEATURES

A single board consisting of
- Insiders (managers, large shareholders)
- Independent (outsider) directors, and
- Grey directors (e.g., partners in the law firm representing the company)
Decisions made by majority vote
Board presided over by chairman

Slide 59

SARBANES-OXLEY (SOX) ACT

After a number of U.S. corporate scandals, e.g., Enron
SOX mandated a number of reforms in U.S. governance systems
- Reporting requirements
- Analyst independence
- Accounting disclosures
Our focus is only on the governance effects of SOX

Slide 60

POST SOX FEATURES OF U.S. BOARDS

Lead directors
- An outside director
- Chairs separate meeting of non-insider directors.
- Frequently has a budget authority for conducting investigations w.o. CEO approval. (in UK, SID (Senior Ind. Director) has similar role)
Executive Session
- Outside directors have special sessions that include only outsiders where management performance is discussed

Slide 61

U. K. BOARD SYSTEM

Principle based regulation-"comply or explain" (Cadbury report reaffirmed in Walker Report)
- Firms need not follow board guidelines but must provide disclosure and explanation of any deviations.
- Guidelines are best practices to which firms should aspire provided by the Combined Code
Best Practices in Combined Code
- At least 50% of board outsiders
- Board must formally evaluate performance of its members and committees
- Remuneration committee all outsiders
- Outsider terms less than 9 years and subject to annual re-election
- CEO should not be chairman
- Publicly disclosed attendance records for meetings

Slide 62

CHANGES SINCE 2007 FINANCIAL CRISIS

Combined code now called "U. K. Code of Corporate Governance"
Principles-based regulation as before
For banks, sets new best practices standards for the role of the risk manager

Slide 63

U.K. GOVERNANCE: SHAREHOLDER POWER

Compared with U.S. and European governance systems, the U.K. system provides shareholders with more direct power over governance
- In UK, all directors must be approved by the majority of shareholders, thus, it is possible to remove directors without proposing alternative directors.
- In U.S. only a plurality of votes is required with the nominee receiving the largest number of votes elected.
- 10% of shareholders can call an Extraordinary General Meeting (EGM), and, with the agreement of 50%+1, remove any or all directors.
- In U.S. (unless the charter otherwise provides) shareholders cannot call EGMS.

Slide 64

TWO-TIER BOARDS

Norm in Germany, Austria, common in Low Countries
Two Separate Boards (non overlapping)
- Insider Board (Management)
  - Makes most operating decisions
- Outsider Board (Supervisory)
  - Controls audit committee
  - Controls compensation committee
  - Hire/fire management
  - Approval required for a few major decisions

Slide 65

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
- Formal institutions
- How governance actually operates
SUMMARY

Slide 66

TRADITIONAL TYPOLOGY OF GOVERNANCE SYSTEMS

Hirschman (1970) develops the classic typology of governance systems:

Anglo-Saxon (AS) model "Exit" model:
- a well-developed stock market with strong investor protection,
- substantial disclosure requirements,
- shareholder activism,
- takeovers.
German-Japanese (GJ) model "Voice" model:
- Trust between management and shareholders.
- Cross-shareholding among firms, and between firms and banks.

Slide 67

HIRSCHMAN: AS MODEL

Managers/Directors: undiversified shareholding by agents with formal control rights over the firm
Blockholders: Large investors whose stake in the firm represents a considerable fraction of their undiversified investment portfolio.
- Private investors
- Founding family (e.g., Walmart and the Walton family)
Retail investors: Many diversified investors each holding a negligible share of the firm's shares

Slide 68

HIRSCHMAN AND AS GOVERNANCE

Managers have incentives to act opportunistically
Outside investors have tiny stakes and are passive: free-rider problem
- if unhappy, then exit
Insider undiversified blockholders, whose wealth is tied up in the firm, are willing to expend resources to check managerial opportunism

Slide 69

THE NEW KID ON THE BLOCK: INSTITUTIONAL UNIVERSAL OWNERS

Around 2000, long after the Voice vs. Exit theory was developed
The growth of indexed-mutual funds, 401k retirement accounts, and pension funds, birthed a new and very important type of share owner: the institutional universal owner

Slide 70

INSTITUTIONAL UNIVERSAL OWNERS

Vary large investment firms in terms of assets under management, e.g. Black Rock, Fidelity
Hold diversified portfolios that nevertheless include large fractions (≈5%) of many firms
Holdings frequently motivated by indexing strategies
Universal owners unlike the traditional investor types: are both
- Large block holders in individual firms
- Diversified holders of large share blocks of many other firms

Slide 71

EXAMPLE: LARGEST SHAREHOLDERS IN TARGET INC.

Largest shareholders in Target (large U.S. retailer) in 2017. (Schmalz, 2018)
Shareholder	Stake (%)
State Street	9.43
BlackRock	8.37
Vanguard	6.73
Franklin Resources	3.80
Dodge & Cox	3.63
Capital Research	3.34
BNY Mellon	1.64

Slide 72

UPDATED TYPOLOGY OF BLOCKHOLDERS

	Insider	Outsider (%)
Diversified		Institutional
Undiversified	Family/Mgrs./Pvt.	Hedge Fund

Slide 73

INSTITUTIONAL UNIVERSAL OWNERS: MODUS OPERANDI

Intervene in governance when firm performance relative to peers is significantly lagging or when governance structures are poor, e.g.,
- Supermajority rules
- Classified boards, etc.
- Poison Pills
- Lack of board diversity
- etc.
Institutional owners' payoffs are determined by assets under management and are not substantially affected by the performance of any one firm in their portfolios
So frequently their proxy voting behaviour is influenced by "optics" rather than ruthless NPV maximization.

Slide 74

HEDGE FUND INVESTORS: MODUS OPERANDI

Hedge fund investors (e.g., Pershing Square Capital Management, William Ackman) aim to improve value drivers (investment, personnel, capital structure)
Hedge Funds aim to
- accumulate undiversified holdings in select target companies
- generate returns by acquiring firm shares and using the resulting leverage over management to
- change operating, governance polices, or induce sale of the firm to an acquirer
- sell holdings for a profit

Slide 75

HEDGE FUNDS: TACTICS

Tactics: Common
- Seek board representation without proxy contest
- Seek board representation through a proxy contest
- Launch proxy contest to replace board
- Make formal shareholder proposals
Tactics: Less common
- Sue the target
- Attempt hostile takeover of the target
- Threaten to sue the target

Slide 76

HEDGE FUNDS VS. UNIVERSAL OWNERS

Institutional universal owners have little to gain from supporting hedge fund campaigns that will make the firm a stronger competitor relative to industry rivals
- The institutional universal owners might have an equally large or larger stake in the industry rivals
Institutional universal owners have little to gain from supporting hedge fund activism campaigns that aimed implement value-add plans that are inconsistent with social or environmental norms even if they are very profitable, e.g.,
- new investments in hydrocarbons
- off-shoring manufacturing
- reducing capital investment in deprived areas

Slide 77

OUTLINE

FACTS
COMPENSATION
TRADITIONAL THEORIES OF MANAGEMENT COMPENSATION
NEWER THEORIES OF MANAGEMENT COMPENSATION
GOVERNANCE
SUMMARY

Slide 78

SUMMARY: COMPENSATION

CEOs compensation packages are high and much more complex than compensation packages for lower level workers
P/A Theory no doubt explains some features of CEO compensation such as the link between pay and performance
However, CEOs do have a rather unique ability to influence their own compensation packages which, no doubt, sometimes lead to skimming
However, skimming does not seem to be a good explanation for overall trends in the managerial labour market like more external hiring, shorter tenure, relative performance evaluation
The newer theories help explain how these trends effect CEO compensation

Slide 79

SUMMARY: GOVERNANCE

The classical Hirschman (1970) Exit (AS) and Voice (GJ) perspective on governance makes a sharp distinction between
- Anglo-Saxon governance systems based on shareholder power and corporate control markets
- and German/Japanese systems where large banks and shareholders exercise cooperative control over firms
The emergence of institutional universal owners in AS countries who, like large banks in the GJ system
- Own significant stakes in many firms
- Have objectives beyond maximizing the value of any single firm
has eroded the distinction between these two systems to some extent

CORPORATE FINANCE VIII: CORPORATE GOVERNANCE

OUTLINE

OUTLINE

FACTS: U.S. CEOS - HOW: SALARY & PERFORMANCE BONUSES FOR U.S. CEOS

HOW: STOCK OPTIONS

HOW: RESTRICTED STOCK AND SEVERANCE PACKAGES

HOW: RELATIVE PERFORMANCE EVALUATION (RPE) PLANS

OUTLINE

FACTS: THE REST OF THE WORLD - HOW ELSEWHERE

CEO COMPENSATION AROUND THE WORLD

OUTLINE

TRADITIONAL THEORIES

OUTLINE

IDEA BEHIND PRINCIPAL/AGENT (P/A) MODEL

A SIMPLE P/A MODEL

ASSUMPTIONS

EFFORT AND CASH FLOWS

MINIMAL COMPENSATION TO ENSURE HIGH EFFORT

EFFORT ASSURING COMPENSATION

VALUE OF HIGH-EFFORT ASSURING COMPENSATION

COMPENSATION FOR HIGH EFFORT

COMPENSATION FOR LOW EFFORT

OPTIMAL POLICY: HIGH OR LOW EFFORT?

OPTIMAL POLICY WITH NO P/A PROBLEM

DISTORTIONS FROM AGENCY PROBLEM

EFFECTS OF AGENCY CONFLICTS

AGENCY COSTS

OBSERVATIONS

OUTLINE

"AGENTS WITH OR WITHOUT PRINCIPALS" (BERTRAND & MULLAINATHAN, 2000)

NUMBER OF HOSTILE BIDS

"AGENTS WITH OR WITHOUT PRINCIPALS" - II

PROBLEMS WITH SKIMMING THEORY: I

PROBLEMS WITH SKIMMING THEORY: II

OUTLINE

NEWER THEORIES

OUTLINE

TOURNAMENT THEORY: IDEA

COSTS OF TOURNAMENT COMPETITION: RISK-TAKING

COSTS OF TOURNAMENT COMPETITION: SABOTAGE

OUTLINE

POSITIVE ASSORTATIVE MATCHING (PAM)

MATCHING: EXAMPLE

MATCHING EQUILIBRIUM

EQUILIBRIUM EQUATIONS: EXAMPLE

GRAPH OF EQUILIBRIUM COMPENSATION

GENERALIZING THESE IDEAS

POSITIVE ASSORTATIVE MATCHING

PAM EQUILIBRIUM

WHAT DOES THIS HAVE TO DO WITH CEO COMPENSATION?

MORAL OF THE STORY

IMPLICATIONS OF PAM

OUTLINE

GOVERNANCE

OUTLINE

CORPORATE BOARDS: WHAT?

FEATURES

FEATURES

SARBANES-OXLEY (SOX) ACT

POST SOX FEATURES OF U.S. BOARDS

U. K. BOARD SYSTEM

CHANGES SINCE 2007 FINANCIAL CRISIS

U.K. GOVERNANCE: SHAREHOLDER POWER

TWO-TIER BOARDS

OUTLINE

TRADITIONAL TYPOLOGY OF GOVERNANCE SYSTEMS

HIRSCHMAN: AS MODEL

HIRSCHMAN AND AS GOVERNANCE

THE NEW KID ON THE BLOCK: INSTITUTIONAL UNIVERSAL OWNERS

INSTITUTIONAL UNIVERSAL OWNERS

EXAMPLE: LARGEST SHAREHOLDERS IN TARGET INC.

UPDATED TYPOLOGY OF BLOCKHOLDERS

INSTITUTIONAL UNIVERSAL OWNERS: MODUS OPERANDI

HEDGE FUND INVESTORS: MODUS OPERANDI

HEDGE FUNDS: TACTICS

HEDGE FUNDS VS. UNIVERSAL OWNERS

OUTLINE

SUMMARY: COMPENSATION

SUMMARY: GOVERNANCE