Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
CHAPTER 10
EXECUTIVE COMPENSATION
10.1 Overview
10.2 Are Incentive Contracts Necessary?
10.3 A Managerial Compensation Plan
10.4 The Theory of Executive Compensation
10.4.1 The Relative Proportions of Net Income and Share Price in Evaluating
Manager Performance
10.4.2 Short–Run and Long–Run Effort
10.4.3 The Role of Risk in Executive Compensation
10.5 Empirical Compensation Research
10.6 The Politics of Executive Compensation
10.7 The Power Theory of Executive Compensation
10.8 The Social Significance of Managerial Labour Markets that Work Well
10.9 Conclusions on Executive Compensation
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
LEARNING OBJECTIVES AND SUGGESTED TEACHING APPROACHES
1. To Establish the Desirability of Incentive Contracts
When people are exposed to agency theory for the first time, a common reaction is to
reject it because they feel that, almost by definition, managers do not need incentives to
want to work hard. I counter this argument by asking the class whether they would work
hard in this course if I was to cancel the final exam. However, in a multi–period context,
the reaction has to be taken seriously. Fama (1980) must have had a similar reaction, and
his argument that market value and reputation considerations will eliminate the shirking
problem is worth outlining.
Fama’s argument can be questioned, however, from both theoretical and empirical
perspectives. In an undergraduate course, I do not pursue the theoretical issues; but the
empirical perspective is worth developing. The first empirical evidence that reputation
effects do not fully eliminate moral hazard (other than casual observation that all large
firms do have incentive contracts) of which I am aware is the paper by Wolfson (1985).
Relevant parts of this paper are described in the text. However, I usually assign the first
part of the paper itself as supplemental reading and work through selected parts of it in
class, since the argument is tortuous. Pages 101–117 of the article, incl., and the text
discussion in Section 10.2 are sufficient for the point to be made that Wolfson’s results
suggest that reputation effects do not completely overcome moral hazard.
I have added a second reference (Bushman, Engel, and Smith (2006)) for anyone who
wishes to pursue this topic further.
2. To Flesh Out Implications of Agency Theory for Executive Compensation
Executive compensation of financial institutions has come under considerable criticism
following the 2007–2008 market meltdowns. Financial institutions are responding with
some compensation changes, as illustrated by the outline of the Royal Bank of Canada
(RBC) Executive compensation plan in Section 10.3. Suggested points to discuss are:
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
373
(i) As predicted by Holmström (1979), the RBC is complex. Of course, part of
this complexity arises because of the multi–period nature of real compensation plans,
whereas Holmström’s agency model spans a single period. Nevertheless, the mixture of
performance measures is consistent with Holmström’s prediction.
(ii) Managers bear risk. This is the other side of the coin from the incentive
effects of giving the manager a share of the payoff. The manager must bear risk so that
working hard is credible to the principle ex ante. Thus, there are minimum stock
ownership guidelines. For example, the CEO should hold stock with a value of 8 times
salary. Also, the long–term incentives plan awards stock options. However, the amount of
risk is controlled by filtering annual performance awards through the Compensation
Committee of the Board, and by positioning total compensation at the median of a peer
group of companies. Also, the use of stock options may help control downside
compensation risk.
(iii) Discussion of the RBC plan, and the proposed changes, could usefully
proceed in conjunction with discussion of the incentive characteristics of stock options,
which have been singled out as one of the reasons for opportunistic manager behaviour
such as that of Enron, and of financial institutions leading up to the 2007–2008 market
meltdowns. Note, in particular, the long term of the ESOs and the relatively slow rate of
vesting. Are these restrictions enough to deter excessive risk taking? In this regard, see
Theory in Practice 10.1.
3. To Understand the Qualities Needed by a Good Performance Measure
The basic qualities needed are precision and sensitivity. Sensitivity is the rate of change
of the performance measure with respect to effort—the more sensitive the measure, the
greater the shift. Precision is the reciprocal of the variance of the density function—the
more precise the measure the less likely it is that random state realization will generate a
performance result greater or less than expected given the manager’s effort. Questions 13
and 19 of Chapter 9 illustrate sensitivity and precision for a 2–point earnings distribution.
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
Instructors may wish to consider the analogy with relevance and reliability, although the
two sets of concepts are quite different. Nevertheless, the need to trade–off is common to
both, when net income is a biased predictor of the payoff. When net income is an
unbiased predictor, the expected value of net income remains equal to the payoff, but
better reporting can still increase precision (Example 9.3).
4. To Evaluate the Role of Net Income in Compensation Plans in Relation to
Stock–Based Performance Measures
The class readily sees the short–run decision horizon that may be induced by basing
manager compensation on the current year’s reported income, and the longer–run horizon
hopefully encouraged by stock–based compensation. Also, share price is more timely in
capturing all the effects of current manager effort, such as R&D. Thus, stock–based
performance measures are more sensitive than net income. However, the class also sees
that stock-based compensation may be more volatile than compensation based on
reported net income. That is, it is less precise since it is affected by economy–wide events
that may have low informativeness about manager effort. It is worth emphasizing that
excess volatility can reduce contracting efficiency – for a risk averse manager, greater
volatility of the performance measure will require higher average pay to maintain
reservation utility, and will discourage the taking of risky projects. However, too little risk
discourages effort. The key is to find a good balance between too much and too little risk
imposed on the manager.
The foregoing suggests that net income competes with stock price as a variable in
compensation plans. I then suggest that both net income and share price (and possibly
additional performance measures, such as attainment of personal goals) are desirable
components of the plan, consistent with Holmström’s (1979) analysis. Also, I bring out that
we can think of manager effort as a two–dimensional variable – call it long–run effort and
short–run effort – and that the relative attention that the manager gives to each dimension
can be controlled by the proportion of current net income–based and stock–based
compensation in the compensation plan. This argument is based on the analysis in
Bushman and Indjejikian (1993).
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
375
Instructors who wish to develop further the concept of multi–dimensional effort can discuss
the concept of non–congruency of a performance measure explained in optional Section
10.4.2. However, this section can be ignored with little loss of continuity.
Having established that there is a role for net income in compensation plans, I then ask
whether the measure of income should be historical cost–based, or current value–based as
per the measurement perspective of Chapters 6 and 7. I argue that historical cost–based
accounting has greater precision, due to the random factors that can affect fair value–
based income but which are low in informativeness about manager effort (e.g., changes in
market prices).
This contracting role for net income is a possible reason why historical cost–based
accounting has such “staying power” in the mixed measurement model, and raises the
question of whether historical cost–based or current value–based income measures are
preferable in compensation plans.
5. To Introduce Empirical Evidence on the Role of Net Income in Compensation
Plans
The predictive ability of the theory of executive compensation needs to be tested
empirically. Beginning with the 1987 paper by Lambert and Larcker, the empirical
research supports the compensation theory, much like the empirical research concerning
investor reaction to accounting information supports the rational investor and market
efficiency theory.
6. To Evaluate Political Aspects of Executive Compensation
Executive compensation has attracted a lot of media and public attention, often focusing
on the question of whether North American managers are overpaid, particularly in relation
to their counterparts in other regions. This question is an excellent one with which to
motivate student interest in the theory. Problems 4, 12, 14, 15, 17, 18, and 19 deal with
various aspects of possible executive over–compensation.
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
376
I also remind the class of the problems of measuring the stock options component of
executive compensation that were discussed in Section 8.6. These problems arise in large
part because the Black/Scholes formula assumes that options are held to maturity
whereas ESOs may be exercised after vesting but prior to maturity. This tends to lower
the ESOs’ fair values relative to Black/Scholes. Recall that, to compensate for this,
accounting standards suggest using the expected time to exercise in Black/Scholes. Also,
it should be noted that since managers are, in effect, forced to hold their ESOs from the
grant date to vesting, this further reduces the ESO value to the manager since in general,
fair value assumes that assets and liabilities can be readily sold if desired. The paper by
Hall and Murphy (2002) brings out the decline in the value of an option to a risk–averse
manager, relative to its Black–Scholes value, as the manager’s ability to sell the option is
restricted. Theory and Practice 10.5 and Problem 13 re Zion’s Bancorporation illustrate
this effect. The main point I make as a result of these considerations is that the value of
ESOs to the manager is likely considerably less than the amounts one sees in the media,
which typically take the number of options exercised times share market value as their
measure of the options compensation.
7. The Power Theory of Executive Compensation
I suggest using this section to bring out the point that just as rational investment and
efficient markets face competing theories, so does the theory of executive compensation.
The power theory argues that managers use their influence in their organizations to
receive more than reservation utility, implying low efficiency of managerial labour markets
and increasing the scope for financial reporting of manager compensation. Discussing the
parallelism here, I hope, helps the students to better understand both theories of reporting
to investors and theories of manager compensation.
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
377
SUGGESTED SOLUTIONS TO QUESTIONS AND PROBLEMS
1. a. The 10% bonus plan is designed to overcome the moral hazard problem
which arises because the shareholders cannot observe management’s effort in
operating the firm. If they received a straight salary, managers may be tempted to
shirk. By giving managers a share in profits (which are informative about effort and
payoff), their tendency to shirk is reduced.
b. Reasons for the Employee Stock Option Plan:
• To strengthen employees’, including senior managements’, incentive (in
addition to the bonus plan) to work hard on the firm’s behalf. Since the
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
378
• To motivate a broader segment of the firm’s workforce – the plan applies to
employees in general, not just officers. The firm may feel that a bonus plan
for all employees may require too much cash. Effort motivation can be
achieved, without direct expenditure of cash, by issuing ESOs. There is an
379
2. Note: Instructors who assign this problem may wish to refer students to Chapter 7,
Note 22, which points out that risk goes both ways. That is, the relevant ex ante
compensation risk measure for a manager who trades–off risk and return is a
measure of the dispersion of compensation, such as its standard deviation.
a. Reasons why it is important to control or reduce the risk imposed on
managers:
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
380
b. Different ways of reducing risk are:
Note: If stock price falls substantially, and for an extended period, the
exercise price of managers’ stock options may be lowered by the
compensation committee of the Board. This possibility further lowers
managers’ risk, but has incentive connotations. See Problems 14 and 15.
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
381
• Earnings management, that is, the ability to choose from a set of
If too much compensation risk is eliminated, the manager may shirk. Agency theory
tells us that if effort is to be motivated the manager must bear some risk. An
efficient contract imposes an optimal amount of risk, not minimal risk.
c. Such requirements are imposed so that managers do not reduce their risk
by selling off their firms’ shares (which are subject to firm–specific risk) and buying
d. Basing managerial compensation solely on share performance may impose
excessive risk on the manager because:
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
382
• Share price tends to be low in precision. Unless offset by higher sensitivity,
the resulting increase in ex ante risk may require higher expected
compensation to maintain reservation utility.
3. No, you should not agree. While cash is received, the amount is less than the
market value of the underlying shares. Consequently, the firm will suffer an
4. The reason why the values of ESOs and restricted stock to a manager are less
than their fair values is that compensation plans generally restrict the manager’s
5. A low pay/performance relationship is expected for large corporations for the
following reasons:
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
• Too much downside risk on a manager reduces compensation contract
6. Firm A’s compensation plan suggests that the precision of net income is high, with
reasonable sensitivity of effort relative to the sensitivity of share price. An example
of an A–type firm is one with the volatility of net income low relative to the volatility
of share price (e.g., low financial and operating leverage), low and stable R&D
385
8. The reason for requiring executives to hold company shares is to further align their
interests with those of shareholders. Also, executives would be motivated to
maintain a longer–run decision horizon.
Note: Executives of financial institutions leading up to the 2007–2008 market
9. Reasons why TD would voluntarily expense its ESOs:
• To show a commitment to transparency, and high quality reported net
$1.076 billion.
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
• TD may have been planning to reduce its use of ESOs in view of the
10. a. If securities markets are efficient, Microsoft’s share price would be
unaffected by the reduction in earnings since the switch generates little effect on
the firm’s cash flows.
Reasons why Microsoft’s stock price may fall include:
• The restricted stock has an exercise price of zero, unlike ESOs. The market
387
• Investors may interpret Microsoft’s switch to restricted stock as an indication
that the firm did not expect its share price to rise to the ESO exercise price
for some time.
b. Reasons why Microsoft’s share price might rise:
c. Microsoft may have wanted to remove market concerns about pump and
dump, and other types of dysfunctional manager behaviour as documented by
Aboody and Kasznik (2000) and Yermack (1997) (see Section 8.6). To the extent
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
388
11. a. The GE compensation plan seems likely to induce a balance of longer–run
effort, for the following reasons:
• The 5–year period before the cash flow and share price targets pay off. This
encourages longer run effort since there will be time for longer–term projects
to pay off. Projects that have a short–run payoff but which may depress
company performance in the longer run (e.g., excessive cost–cutting) will be
discouraged.
b. Dysfunctional effects of too much compensation risk include:
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
389
next period, so that cash flow is less informative about effort than net income,
which would accrue the amount receivable. Similarly, cash payments may be
deferred if the firm runs short of cash, whereas the related expenses would be
accrued as accounts payable.
Operating cash flows may also be less sensitive than net income. For example, a
Scott, Financial Accounting Theory, 7th Edition Instructor’s Solutions Manual Chapter 10
390
Eliminating unusual events from the operating cash flow performance measure
seems to be a risk–reducing device. Such events are often hard to predict and are
large in amount. Eliminating them reduces risk by increasing the precision of the
if the number of restricted shares to be granted depends on share price
performance.
• Restrictions on disposal during the vesting period reduce the temptation to
pump share price during that period. This is especially so if the vesting
period is long—5 years in the case of GE. To the extent restricted stock
12. a. The argument for disclosure of compensation information of senior
executives is to motivate manager effort and to enable managerial labour markets