What Determines Stock Option Contract Designs

What determines stock option contract design?
$
Eva Liljeblom, Daniel Pasternack
n
, Matts Rosenberg
Hanken School of Economics and Business Administration, P.O. Box 479, 00101 Helsinki, Finland
a r t i c l e i n f o
Article history:
Received 19 January 2010
Received in revised form
14 December 2010
Accepted 15 February 2011
JEL classication:
G30
G32
J33
Keywords:
Stock option contract design
Optimal contracting
Agency cost
a b s t r a c t
We analyze the factors that drive exercise price policy for executive option plans
(ESOPs) and their scope in a country where rms are not subject to the tax and
accounting considerations that seem to have led to the dominance of at-the-money
options in the US Our unbounded data for Finland provide us with an excellent
opportunity to investigate whether contract design is consistent with compensation
theory. Our ndings are largely consistent with predictions from the optimal contract-
ing literature. The size of the plan is negatively related to Tobins Q and rm size and
positively related to proxies for monitoring costs, which also inuence the probability
of launching premium ESOPs. Our results also show that the premium (out-of-the-
moneyness) is negatively related to prior stock returns and cash ow-to-assets, which
may be an indication of high-water mark contracting, or alternatively, of managerial
power. Finally, we also nd some support for a positive relation between the premium
and the length of the vesting period when maturity is xed, which indicates an effort to
keep the incentives for management from falling over time.
& 2011 Elsevier B.V. All rights reserved.
1. Introduction
Most empirical studies of equity-based compensation
concentrate on the incentives they provide (e.g., Yermack,
1995). Incentives are typically measured by some (abso-
lute or relative) measure of pay-to-performance sensiti-
vity. For executive options, the pay-to-performance sen-
sitivity is a function of the option delta (i.e., related to the
exercise price in relation to the stock price as well as, e.g.,
stock volatility) and the grant size [see, e.g., the option-
specic pay-performance sensitivity of Hall and Murphy
(2000, 2002)]. Hall and Murphy (2000) consider the
exercise-price policy as perhaps the most central design
issue regarding executive options. Paradoxically, the
variation of the key variable, the exercise price at the
grant date, has been rather limited in the US, where most
executive options have been granted near at-the-money.
1
Hall and Murphy (2000, 2002) provide a rationale for this
phenomenon using elements related to managerial risk
aversion and elements from US tax and accounting con-
siderations.
2
Additional arguments like these are needed
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/jfec
Journal of Financial Economics
0304-405X/$ - see front matter & 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jneco.2011.02.021
$
We are greatly indebted to valuable comments by Professor Li Jin.
We also thank Tom Berglund, Seppo Ik aheimo, Esa Jokivuolle, Anders
L ound, Darius Miller, Vesa Puttonen, the participants of the FPPE
workshop on Capital Markets & Financial Economics, and the partici-
pants of the GSFFA Joint Finance Research Seminar for valuable com-
ments and suggestions.
n
Corresponding author. Tel.: 358 505693416.
E-mail address: daniel.pasternack@hanken. (D. Pasternack).
1
For example, Murphy (1999) documents that 95% of the chief
executive ofcer (CEO) stock options of 1,000 large US rms in 1992
were granted at-the-money. More recent results indicate that this still is
the case; e.g., Banerjee, Gatchev and Noe (2008) report that in 2005,
99.92% of new option grants to CEOs were issued at-the-money. For
Europe, Sauer and Sautner (2008) nd that when underwater options are
repriced, the new exercise price is set so that the median (mean) ratio of
the new exercise price to the reprising day stock price is 1.00 (1.21).
Their results thus indicate that in Europe as well, setting option exercise
prices near at-the-money is favored. In, e.g., Australia, however, a larger
variation in exercise prices can be found; see, e.g., Rosser and Canil
(2004).
2
Their main explanation is based on an analysis of the optimal pay-
to-performance sensitivities given risk-averse managers who may have
a large proportion of their wealth in company stock or options. They nd
that the incentive-maximizing exercise price range is relatively at and
Journal of Financial Economics ] (]]]]) ]]]]]]
Please cite this article as: Liljeblom, E., et al., What determines stock option contract design? Journal of Financial
Economics (2011), doi:10.1016/j.jneco.2011.02.021
because, without them, the predominance of at-the-
money options would partly remain a puzzle given the
atness of the incentive-maximizing exercise price range.
We provide unique evidence of the determinants of
stock option design using data for a market where
exercise price setting is not restricted by tax or account-
ing considerations. Such a setting allows us to investigate
the factors that drive the key variable, exercise price,
which varies little in the US Recent contributions by,
for example, Dittmann and Yu (2009) and Palmon and
Venezia (2009) suggest that options should be in-the-
money (at least in the absence of US tax considerations).
Our sample, where the setting of the exercise price is
unbounded by institutional restrictions, serves as an
excellent experimental laboratory for such theoretical
predictions. In addition to exercise price, we examine
another important design attribute, the scope of the plan
[for other important features of stock option programs,
such as vesting schedules and contract maturity, see, e.g.,
Rosenberg (2003)]. Finally, we also study the effect of the
vesting period on the exercise price. Bebchuk, Fried, and
Walker (2002) notice that little attention is given to the
fact that rms (in the US) use the same exercise price for
options regardless of the vesting period. Our tranche-
specic stock option data for Finland allow us to study
this issue, in a setting where exercise price setting is
relatively free of restrictions as compared with the US.
We nd that the ratio of the exercise price to the stock
price on the granting date (the out-of-the-moneyness of
the option) is negatively related to the prior stock return.
This result may be interpreted as a shareholder response
to a poor prior stock price performance by requiring a
higher subsequent stock price appreciation to reward
managers (high-water mark contracting). Alternatively,
the result suggests that managers have more negotiation
power regarding the design of compensation in rms with
greater prior stock price performance. Our additional tests
do not allow us to strongly distinguish between these two
alternatives. Moreover, our results indicate that invest-
ment intensity, cash ow, and monitoring costs are
associated with the likelihood of granting premium
(out-of-the-money) stock options.
Our results show that the scope (measured as the
fraction of equity obtained upon the exercise of all
granted stock options) is negatively related to Tobins Q.
If Tobins Q is treated as a measure of rm performance,
this relation would suggest that poorly performing rms
grant larger stock option plans with greater scope. In
addition, we nd that the scope of stock option plans
increases simultaneously with proxies for monitoring
costs, which is consistent with the traditional principal-
agent theory that greater monitoring cost/difculty
should be positively related to the amount of equity-
based compensation (Holmstrom, 1979; Demsetz and
Lehn, 1985; Milgrom and Roberts, 1992). We also nd
that the scope of the stock option plan is greater in broad-
based plans and in plans with dividend protection.
Finally, we nd that when institutional restrictions
concerning exercise price setting are absent, there is a
signicant positive relation between the out-of-the-money-
ness of the option tranches and the length of their vesting
period. This nding is contrary to the view of Bebchuk,
Fried, and Walker (2002) that managerial power leads to the
same exercise prices ( ya royalty on the passage of time,
as Buffett put it
3
). Instead, our results suggests that when
both the size of the compensation program and the exercise
price are allowed to vary, rms set higher exercise prices for
options with longer vesting periods to keep the incentive
effect from diminishing over time.
We contribute to the relatively thin literature (as noted
by Dahiya and Yermack, 2008) on inter-rm differences in
the design and implementation of executive stock option
plans. There are accepted theoretical arguments for using
different design elements in different types of rms, for
example, due to differences in monitoring costs, but very
little empirical evidence exists on their actual validity.
Furthermore, while there has been little variation in the
exercise price in the US and thus little focus on it as a
variable, the introduction of FAS 123(R) in 2005 may
change this situation. By introducing mandatory options
expensing irrespectively of the in-the-moneyness of the
option, FAS 123(R) has ruled out some of the earlier
obstacles to options other than xed-plan at-the-money
options. As a result, evidence from another market, where
options with varied exercise prices have been subject to
equal treatment in tax and accounting rules, may be of
interest with respect to the potential future development
of other markets.
The paper is organized as follows. Section 2 provides
background information on option grants in Finland and
on the comparability of Finland to other markets. Section
3 presents the hypotheses and the research design used in
the study. Sample selection and data characteristics are
discussed in Section 4. Section 5 reports the empirical
results, and Section 6 concludes the paper.
2. Institutional details for Finland
2.1. Governance and executive compensation in Finland
Although Finland is a small country compared to the
US, we believe that analyzing the setting of exercise prices
in Finland can provide insight into the factors that might
drive exercise price setting in other markets, such as
(footnote continued)
mostly surrounds the at-the-money point. However, e.g., Hall and
Murphy (2000) believe that US tax and accounting rules can explain
the paucity of discount (in-the-money) options (but not the paucity of
premium options). In the US, in-the-money stock options have not been
regarded as performance-based compensation under Section 162(m)
of the Internal Revenue Code and thus do not constitute a tax-deductible
cost if an executives total non-performance-based compensation
exceeds $1 million per year. In addition, prior to 2005, stock options
that were granted in-the-money had to be included in the income
statement as a compensation expense, whereas xed-plan options
granted at-the-money and out-of-the-money did not have to be. With
the introduction of Financial Accounting Standard (FAS) 123(R) came a
mandatory option expensing for nancial statements with scal years
beginning after June 15, 2005, which may inuence the design of newer
executive option programs (Cuny, Martin, and Puthenpurackal, 2009).
3
Speech held at Berkshire Hathaway annual meeting 2003.
E. Liljeblom et al. / Journal of Financial Economics ] (]]]]) ]]]]]] 2
the US, when the institutional factors that drive towards
at-the-money options are no longer present. Compared
with continental European countries, Finland is in many
ways similar to the US in terms of corporate governance,
business culture, and shareholder orientation. When
Aggarwal, Erel, Williamson, and Stultz (2009) compare
foreign rms to US rms in terms of their level of
corporate governance (CG), they nd that the top four
CG countries are Canada (a CG score of 68%), the US (59%),
Finland (56%), and the U.K. (55%).
4
Also in terms of ownership concentration, Finland
represents a middle case between the typical European
pattern of concentrated ownership and the ownership
structure in the US.
5
With decreasing ownership concen-
tration, and increased pension savings, institutional own-
ers have become increasingly important in the Finnish
market [see, e.g., Bhattacharya and Graham (2009), who
study institutional owners in Finland]. Many institutional
investors in Finnish stocks are foreign funds. Several
authors argue that the high degree of internationalization
has driven Finnish CG and business culture towards more
a competitive culture where shareholder value is given an
increasingly high priority.
6
During our study period, the compensation mix for top
executives in Finnish-listed rms typically includes a base
salary, a bonus, and executive options. Additional retire-
ment (pension) benets are also common. Stock-based
compensation has gained in popularity since 2002, i.e.,
after the period of the technology stock-market crash,
when the popularity of option plans began to fade. In
2005, stock-based compensation became slightly more
common than executive options (Ik aheimo, Kontu,
Kostiander, Tainio, and Uusitalo, 2007). In an international
comparison [Ik aheimo, Kontu, Kostiander, Tainio, and
Uusitalo (2007), using data from 2005 for Finland, and
data from 2006 for other markets], the compensation
structure in Finland was found to be similar to that of
other European countries such as the U.K., Italy, France,
and Sweden. Base salary counts for about 42% of total
compensation, variable payment (bonus, options and/or
stock-based compensation) for 35%, and pension and
other benets for the remaining 23%. In the US, the
corresponding percentage for base salary is only 27%.
Despite some differences in the compensation mix,
executive compensation practices in Finland are similar to
those in the US in terms of their stock price sensitivity.
Makinen (2007) nds that in Finland, the average compen-
sation of CEOs increased substantially over time, and that
this change, and especially when measured in terms of total
compensation, is highly related to changes in stock market
measures of rm performance. His estimates of semi-
elasticities of CEO salary and bonus with respect to stock
returns are between 0.09 and 0.28 for total compensation in
Finland, values that are in line with those of Rosen (1992),
who nds that the estimated semi-elasticity of CEO
salaries and bonuses with respect to stock returns is in the
0.100.15 range for the US and the U.K. M akinens (2007)
ndings (for a more recent time period, 1996 to 2002) also
suggest that CEO pay-for-rm size elasticity does not
deviate substantially from 0.3, the number reported by
Rosen (1992). Interestingly, M akinen also nds that the
share of foreign ownership is positively and signicantly
associated with the level of CEO compensation.
2.2. Characteristics for Finnish executive option plans
The rst executive stock options were introduced in
Finland in the late 1980s, and all option plans were issued
as bonds with warrants, until pure employee stock
options were allowed in the amendments to the Compa-
nies Act in 1997. However, options are still sometimes
introduced as bonds with warrants, typically when the
target group is broad-based. In one of four cases (35 out of
141 during our study period), option plans are granted to
a broad group of employees. All other grants are to top
executives (CEO and top management) only.
The typical time-to-maturity and vesting schedule of a
stock option plan in Finland corresponds to the design of
similar contracts in the US (see, e.g., Murphy, 1999). The
average time-to-maturity of the options in our sample is
6.16 years and they become vested (i.e., exercisable) in one
to ve portions (tranches) over a period of about two to ve
years from the date of the grant. This could mean, for
example, that in each of the last three years before the
options expire, one-third of the options become exercisable.
Prior to 1993, options in Finland were taxed in a
similar manner to stocks issued to personnel. If the price
at the date of issuance was more than 15% below the
current market value, the difference constituted a taxable
income, and it was taxed according to a progressive
income tax rate. In 1993, a major tax reform distinguished
income (taxed at a progressive tax rate, which could reach
a maximum of 60% at the margin) from capital gains
(taxed at a lower at rate). Stock options were treated as
capital gains and taxed at a at rate of 25%. However,
4
Aggarwal, Erel, Williamson and Stultz (2009) also nd that only
15 countries have at least one rm that has better governance than its
matching US rm. There are only three countries with more than ve
rms that have a positive index gap, meaning that such rms invest
more in governance than their US counterpart: Canada, Finland, and the
U.K. In the case of Finland, 29.2% of the rms have a positive gap while
for the U.K., the percentage is 24.4%. An example of a specic CG
component where Finland ranked high was also the low level of
staggered boards. They document that four countries (Canada, Denmark,
Finland, and Sweden) have a much higher proportion of rms without a
staggered board as compared to the US.
5
Korkeam aki, Liljeblom, and Pastenack (2010) report that less than
20% of the rms have a block holder owning more than 50% of the shares
in Finland, whereas the evidence in Barca and Becht (2001) indicates
that the percentage is more than 50% in, e.g., Austria, Belgium, Germany,
and Italy. There is also more cross-sectional variation in ownership
concentration in Finland as compared to Continental Europe, from very
widely held rms such as Nokia to rms with a more concentrated
ownership structure.
6
See, e.g., Tainio and Lilja (2003) and Yl a-Anttila, Ali-Yrkk o, and
Nyberg (2004). In the latter, survey results are reported, showing the
high importance of protability, and the increasing importance of
shareholder value, among the goals of large Finnish companies. In
Finland, between 40% to 60% of the stock market is owned by foreign
investors. Yl a-Anttila, Ali-Yrkk o, and Nyberg (2004) report that Finland
ranked as number two among 16 European countries as a cross-border
merger target, and that Finnish industrial corporations are highly
internationalized: amongst the ten largest corporations, as much as
80% of total revenues come from foreign sales and over 60% of the
production and the personnel is located in foreign units.
from September 1994 onwards, option income was trea-
ted as salary and taxed at the higher income tax rate.
Income is calculated as the difference between the market
price and the exercise price at the time of exercise, or
alternatively, using the selling price of the option. This tax
treatment has led to the common practice of selling the
shares obtained through the option exercise almost
immediately after the exercise to ensure liquidity to pay
for heavy income tax. No tax is carried at the grant date.
The option-granting rm has to pay a social security
expense of approximately 5% of the income received by
the option holders from the options, and this expense is
deductible from corporate tax. However, the cost of
granting options is not deductible for the rm. In this
sense, the Finnish tax setting resembles that of Canada
(see, e.g., Klassen and Mawani, 2000), where employers
are disallowed tax deductions and income statement
deductions related to option compensation.
3. Hypothesis development and research design
Compensation theory suggests that the principals
ability to observe the agents actions determines the form
of compensation. If the appropriate actions are known
and observable, the optimal incentive contract pays the
agent (manager) a xed salary and penalizes her for
suboptimal behavior. However, if managerial actions are
at least partly unobservable, tying managerial compensa-
tion to productive outcomes (such as rm value) is
necessary to induce the manager to behave optimally
(Holmstrom, 1979).
The ability of shareholders to observe and evaluate the
actions undertaken by managers is affected by the com-
plexity of managerial tasks. When the decisions required
from managers are more complex, it is more difcult for
shareholders to monitor these decisions. Consequently,
theory suggests that proxies for monitoring cost, such as
rm size, asset complexity, and growth opportunities,
should be related to the amount of the equity-based
compensation and to the power of the incentives that the
compensation provides (see, e.g., Demsetz and Lehn, 1985;
Milgrom and Roberts, 1992). This prediction has gained
strong support in empirical work.
7
More recently, however,
the optimal contracting literature has been questioned by
authors who suggest that the form and design of executive
compensation is determined by managerial power and the
extraction of rents (Bebchuk, Fried, and Walker, 2002). To
contrast and test these theories, we employ as determi-
nants variables from both the optimal contracting litera-
ture and the managerial power literature.
The following subsections discuss the hypothesized
relations between rm characteristics and stock option
contract design. First, in Sections 3.1 and 3.2, we discuss
the expected relations between rm attributes and the
scope of stock option plans or the exercise price of stock
options. Second, in Section 3.3, we present and motivate
some additional tests.
3.1. The scope of stock option plans
Our dependent variable is the relative size of the stock
option plan. We rst measure this variable by the stock option
overhang (dilution), that is, the fraction of equity obtained
upon the exercise of all granted stock options. Second, we
measure the scope as the BS value to MV of equity, that is, the
Black-Scholes value of the plan divided by the market value
of equity at the grant date.
8
See Appendix A for a detailed
description of the calculation of the variables.
Under the optimal contracting hypothesis, the rm
(the board/compensation committee and, ultimately, the
shareholders) designs compensation arrangements exclu-
sively for the purpose of alleviating the agency problem
between shareholders and managers (see, e.g., Ross, 1973;
Holmstrom, 1979; Grossman and Hart, 1983; Milgrom
and Roberts, 1992). Hence, under this approach, we
expect a negative relation between CEO ownership and
the scope of stock option plans because direct share
ownership may be treated as a substitute for stock option
compensation. In contrast, under the managerial rent
extraction hypothesis of Bebchuk, Fried, and Walker
(2002), part of the agency problem is that executives
use their compensation to provide themselves with rents
(i.e., payments in excess of those under a contract that
maximizes shareholder value). Because CEO ownership
measures the direct voting power of the CEO, it is
plausible to expect a positive relation between CEO own-
ership and the scope of stock option plans under the
managerial rent extraction hypothesis.
Based on two arguments, ownership control can be
negatively related to the scope of the stock option plans.
First, if concentrated ownership leads to more efcient
monitoring of the management, the need for equity-based
compensation should be reduced. Furthermore, owner-
ship concentration is assumed to be inversely related to
managerial power and should therefore reduce possibili-
ties for managerial rent extraction. Because in Finland
state ownership is high in some rms and because the
state as an owner can potentially monitor in a different
manner than other owners, we measure non-state and
state ownership separately. We expect Non-state owner-
ship control to be negatively related to the scope of the
stock option plans.
Among the non-state holders, institutional owners can
be viewed as more professional shareholders. Professional
7
For example, Baker and Hall (2004) and Himmelberg, Hubbard,
and Palia (1999) show that the value of equity incentives increases at a
decreasing rate with rm size. Furthermore, the ndings of Smith and
Watts (1992), Gaver and Gaver (1993), Mehran (1995), Guay (1999),
Himmelberg, Hubbard, and Palia (1999), and Palia (2001) lend support
to a positive relationship between growth or investment opportunities
and equity incentives.
8
We calculate the value of stock options using the model of Merton
(1973), i.e., modied for dividend payments, for stock options not
protected against dividend payments. For stock option grants speci-
cally protected against dividend payments, we employ standard Black
and Scholes (1973) methodology. Further, to adjust for the fact that
managers may not be able to fully hedge their holdings (a central
assumption in Black-Scholes valuation), and that the rm may take this
into account when designing their option programs, we also recalculate
the Black-Scholes values using the method suggested by Meulbroek
(2001) for a fully undiversied manager.
shareholders are expected to have an articulated interest
in developing corporate governance mechanisms, which
may result in a positive relation between Institutional
ownership and the scope of the stock option plan. In
contrast, managerial power is expected to be decreasing
in institutional ownership, thus supporting a negative
relation for this variable.
An interesting institutional detail in Finland (unavail-
able for examination in the US) is State ownership. In
international studies, state ownership is often considered
as a weak form of governance [see, e.g., Shleifer and
Vishny (1997) for a discussion of problems with state
ownership]. In Finland, fear of loose governance in state-
controlled rms has led to the creation of government
ownership units that manage the role of the state as a
shareholder. Empirical studies
9
indicate that compensa-
tion in Finnish rms with state ownership does not
exceed compensation in other Finnish rms, and that
option programs in Finnish state-owned rms are more
restrictive. The likely cause of these features is high public
scrutiny in Finland concerning top management salaries,
which are disclosed through public tax records and
typically considered to be too high by the media and the
public. This may lead the government to enact policies to
ensure that state-owned rms are not exceptional to
public expectations in this respect. Hence, we expect a
negative relation between state ownership and the scope
of the stock option plan in our data set.
The expected sign for rm size is more ambiguous. One
can approach this question both from the agency (mon-
itoring) perspective and from the perspective of optimal
compensation theory (free of agency costs). From the
agency point of view, Jensen and Meckling (1976) argue
that large rms are more difcult (costly) to monitor,
which would motivate greater equity incentives in large
rms. Also, e.g., Demsetz and Lehn (1985) hypothesize
that the level of managerial equity holdings should be
greater in larger rms. However, other authors take
different views on this relation. Schaefer (1998) reports
an inverse relation between pay-performance sensitivity
(measured on a dollar-to-dollar basis) and rm size.
Because grant size (together with delta) is a component of
pay-performance sensitivity, this result might imply a
negative relation between the scope of option plans and
rm size. Baker and Hall (2004) obtain results in line with
Schaefers (using a similar measure), but they also show
that when changes in the CEOs wealth are related to
percentage changes in rm value (instead of absolute
changes), the relation to size is strongly positive.
10
They also
estimate the elasticity of the CEOs marginal product with
respect to rm size and obtain an estimate of about 0.4; that
is, they nd that CEO marginal products rise strongly, but
not proportionally/not linearily in absolute terms with rm
size. In our model, we use a proportionate (relative) option
plan scope variable (relative to the equity of the rm). If CEO
productivity rises proportionally with rm size and equity
(given unchanged leverage) and if compensation follows
proportionally, our (relative) measure of scope will remain
constant and therefore be independent of any measure of
rm size. However, if CEO productivity rises less than
proportionately with rm size, as argued by Baker and
Hall (2004), a negative relation will follow.
Theory and prior empirical evidence suggests that
equity-based compensation is positively related to rm
growth opportunities. This suggests a positive relation
between the scope of stock option plans and both Tobins
Q and investment intensity, which is measured as Invest-
ment-to-capital. However, several prior studies (see, e.g.,
Morck, Shleifer, and Vishny, 1988) employ Tobins Q as a
measure of rm performance. If one assumes that the
expected marginal benets of equity-based compensation
are decreasing in rm performance (if, on average, better-
performing rms are already closer to the optimal level of
managerial compensation, and enjoy the benets of close to
maximal managerial effort), then the shareholders of well-
performing rms may be less willing to provide additional
large equity incentives because their costs, given risk-averse
managers already owning rm stock, may exceed the
benets.
11
This would result in a negative relation between
scope on the one hand and Tobins Q and protability
(measured as the ratio of Cash ow- to-assets) on the other.
We also include a measure of nancial leverage (Long-
term debt-to-assets) and expect a negative relation between
nancial leverage and the scope of stock option plans.
Jensens (1986, 1993) arguments imply that the disciplinary
role of debt may reduce agency costs, in which case rms
with high nancial leverage have a reduced need for equity
incentives as a control mechanism.
12
We include the
Capital-to-sales ratio and Firm focus to measure monitoring
complexity and expect a negative relation between both
variables and the scope of stock option plans.
13
In addition,
9
See, e.g., Hansson, Liljeblom, L ound, Maury, Pasternack, and
Rosenberg (2002), who study option programs in listed rms from
1989 to 2000, and Ik aheimo, Kontu, Kostiander, Tainio and Uusitalo
(2007), who use compensation and interview data for 2001 to 2005.
Hansson, Liljeblom, L ound, Maury, Pasternack, and Rosenberg (2002)
nd that the option programs in state-controlled rms were smaller in
scope and with regards to the incentives given by them. Ik aheimo,
Kontu, Kostiander, Tainio, and Uusitalo (2007) in turn conclude that the
compensation structure in Finnish rms with state ownership was
similar to that in other Finnish rms. Controlling for sector and year,
the ability of state-owned rms to generate shareholder value was very
similar to that of other Finnish rms, although somewhat more stable.
10
The key question is how a CEOs productivity is related to rm
value; it may be through a constant monetary effect irrespective of rm
size or on a proportional basis. This discussion is related to that in
Holmstroms (1992) two models of pay sensitivity, of which he prefers
the proportionate one, partly due to its better t with the data.
11
Of course, well-performing rms also need to periodically renew
their incentives, which at least partly dampens this argumentation.
12
Furthermore, John and John (1993) analyze the relation between
rms compensation policies and capital structure and predict that highly
levered rms will provide less equity incentives to motivate optimal choices
regarding managerial risk. The intuition of this argument is straightforward,
i.e., if managers have strong incentives to increase the value of equity,
creditors will demand higher risk premia for providing capital (debt) for fear
that managers will pursue high-risk strategies, transferring wealth from
creditors to shareholders (the asset substitution problem).
13
Empirical work has shown that rm diversication destroys value
and results in the well-known diversication discount. The dominant
part of the explanations for the diversication discount suggests that
agency costs can be expected to be higher in diversied rms, which in
we further decompose the variable measuring rm focus
into Mature industry and Growth industry indicators because
we expect to nd differences in the use of equity incentives
between these groups. See Table 2, Panel C for details
concerning which industries are classied as mature vs.
growth industries.
Firm risk is included to measure monitoring difculty
as a result of noisiness in the rms operating environ-
ment (Demsetz and Lehn, 1985) and of the extent to
which risk-averse managers can be incentivized via stock
and option holdings (see, e.g., Himmelberg, Hubbard, and
Palia, 1999). The expected relation between rm risk and
the scope of stock option plans is thus subject to ambi-
guity. In empirical work, Jin (2002) nds that executive
pay becomes less sensitive to performance as unsyste-
matic risk increases but nds no relation between sys-
tematic risk and pay-for-performance. We thus also
examine whether the type of risk affects the relation by
decomposing Total risk into Systematic risk and Unsyste-
matic risk components.
We also include indicator variables for a Prior plan in
effect and a Broad-based plan. We expect a smaller scope
of stock option plans if a prior plan is in effect. A broad-
based plan is expected to have a greater scope because the
target group of the plan is assumed to be larger. Further-
more, we include a dummy for a Dividend-protected plan
and performance-vested/indexed plans (PV/indexed plan)
to control for the effect of unconventional contract design
features. Finally, year indicator variables are included to
control for time effects.
Based on the above discussion, our model for the
determinants of the scope of the option plan is:
Scope of stock option plan f CEO ownership 7,
Non-stateownership control, Institutionalownership 7,
State ownership , Firm size , Tobins Q 7,
Investment to capital ,Cash flow to assets ,
Long-term debt to assets , Capital to sales ,
Firm focus , Risk as Total risk, Systematic risk,
or Unsystematic risk 7, Prior plan in effect ,
Broad-based plan , Dividend-protected plan ?,
PV=indexed plan ?: 1
Appendix B provides a summary of our expected signs
as well as the categories of theories from which they are
derived.
3.2. The exercise price of stock options
Determinants of the second main design attribute of
stock option plans, the setting of the exercise price, are
investigated by a similar model as in the case of the scope
of stock option plans. The empirical model is thus as
follows:
Stock option premiumf CEO ownership ,
Non-stateownership control, Institutional ownership
, State ownership , Firm size ?, Tobins Q,
Investment to capital, Cash flow to
assets , Longterm debt to assets ,
Prior stock return, Capital to sales,
Firm focus , Risk as Total risk, Systematic risk,
or Unsystematic risk 7, Prior plan in effect ?,
Broad-based plan ?, Dividend-protected plan,
PV=indexed plan : 2
Again, Appendix B provides a summary of our
expected signs and the categories of the theories from
which they are derived. Our dependent variable, the stock
option premium (out-of-the-moneyness), is dened as
[(XS)/S], where X corresponds to the exercise price of
the option, and where S is the stock price at the grant
date. We use two different specications of the dependent
variable: (i) the First tranche premium, which is calculated
as the out-of-the-moneyness of the options belonging to
the rst tranche in the corresponding option plan, and (ii)
the Weighted average premium, which utilizes information
on the characteristics of the total option plan. The weights
used in the calculation correspond to the ratio of the
number of shares obtainable upon exercise in each
individual stock option tranche divided by the total
number of shares obtainable upon exercise of all stock
options. See Appendix A for a detailed description of the
procedure used to calculate the stock option premium.
In Eq. (2) for the premium, most of the explanatory
variables are the same as in Eq. (1) and often have similar
interpretations in terms of the effects they proxy. How-
ever, because the dependent variables in (1) and (2) are
opposite in the sense that a larger scope is related to a
higher level of compensation, while a higher premium (a
higher out-of-the-moneyness) means a lower option and
compensation value, the expected signs for the explana-
tory variables are typically opposite to those in Eq. (1). We
therefore expect, for example, that greater managerial
power (higher CEO ownership, and, correspondingly, lower
Non-state ownership control, Institutional ownership, and
State ownership) is negatively related to the stock option
premium. A greater degree of power enjoyed by managers
may lead to more potential benets that they can extract
at the expense of shareholders.
Firm size has an ambiguous expected impact on option
premium from a theoretical point of view but is included
as a control variable. Tobins Q and rm protability (Cash
ow-to-assets) have an expected negative relation with
the stock option premium because we expect that rms
(shareholders) with lower protability (reected in the
stock price) require greater stock price appreciation to
award managers. Further, we expect investment intensity
(Investment-to-capital) and leverage (Long-term debt-to-
assets) to be negatively related to the stock option
premium. Choe (2003) shows that, all other things being
equal, the optimal exercise price of stock options is
(footnote continued)
turn could be explained by the fact that the monitoring of the managers
effort is more costly in diversied rms compared to focused rms. See,
e.g., Lamont and Polk (2001) for a review of the causes and conse-
quences of diversication.
decreasing in the riskiness of the rms desired invest-
ment policy and in the degree of nancial leverage. More
specically, Choe (2003) argues that greater nancial
leverage reduces the exercise value of stock options,
which leads to the need to perform a downward adjust-
ment on the exercise price, if other components of
executive compensation are xed. In other words, higher
leverage encourages managers to take more risk (asset
substitution), which needs to be mitigated through lower
exercise prices of stock options.
We expect Prior stock return to be inversely related to the
stock option premium. Two alternative hypotheses could
produce such a relation. First, rms (shareholders) might
respond to poor prior stock price performance by requiring
a higher corresponding stock price appreciation to reward
managers. This would be in line with the idea that owners
have a high-water mark contract in mind, i.e., after falling
stock prices, they expect a recovery up to a certain point
before rewarding managers for the additional upside.
14
This
explanation would be in line with optimal contracting. An
alternative explanation, following the managerial rent
extraction hypothesis, is that the ability of the manager to
negotiate favorable compensation arrangements should
increase following a stock price appreciation. If Tobins Q
(which also is dependent on stock price) is viewed as a
proxy for protability, the same negative relation with the
stock option premium as for Prior stock return might be
expected, with the same two alternative arguments as
above. Additionally, when Cash ow-to-assets proxy prot-
ability, the managerial rent extraction hypothesis may
apply, which would indicate a negative expected sign
because the managerial power to negotiate favorable con-
tracts might be higher in more protable rms.
The Capital-to-sales ratio, Firm focus, and rm risk are
used to proxy monitoring difculty/costs. Yermack (1997)
argues that rms may want to award discount stock
options with in-the-money exercise prices to achieve a
larger pay-for-performance sensitivity than provided by
at-the-money, fair market value stock options. On a
share-for-share basis, discount stock options provide
higher pay-for-performance sensitivities than do fair
market value stock options (Lambert, Larcker, and
Verrecchia, 1991). Furthermore, this result may be gen-
eralized regarding the exercise price of the stock options,
that is, a higher out-of-the-moneyness of the stock option
indicates a lower degree of pay for performance, all other
things being equal. Consistent with principal-agent the-
ory, we expect that rms with a higher degree of mon-
itoring costs design compensation structures to provide
greater pay-for-performance. The relation between the
stock option premium and both the capital-to-sales ratio
and rm focus (which, when higher, proxy lower mon-
itoring costs) is expected to be positive, whereas the
relation between the premium and rm risk (total risk,
systematic risk, and unsystematic risk) is, as before,
subject to ambiguity since risk may proxy both monitor-
ing difculty as well as the extent to which risk-averse
managers can be incentivized. As mentioned previously,
Jin (2002) nds a negative relation between the level of
unsystematic risk and pay-for-performance but fails to
nd a signicant relation between systematic risk and
pay-for-performance. For the option premium, there is an
additional possible explanation, also under optimal con-
tracting: high-water mark features may be included
especially in high volatility industries. That would indi-
cate a positive relation between rm risk and the option
premium.
Further, in a similar manner as for the scope of stock
option plans in Eq. (1), we examine whether the premium
differs for rms operating in Mature and Growth indus-
tries. Furthermore, we control for the existence of a Prior
plan in effect, whether the plan is a Broad-based plan, a
Dividend-protected plan, or a performance-vested/indexed
plan (PV/indexed plan). The rst two variables have
ambiguous expected relations with the stock option
premium. For dividend-protected stock options we expect
a higher premium, whereas for performance-vested/
indexed stock options we expect a lower premium. This
is because the expected value of the stock option is higher
in the case of dividend protection and lower in the case of
performance-vesting/indexing, all other things being
equal. As for the scope of stock option plans, we control
for time effects by including year indicator variables.
3.3. Additional tests regarding scope and exercise prices
The empirical analysis in this paper concentrates on
the two main design attributes of stock option plans,
namely, the relative size of stock option plans (scope), and
the exercise price. These attributes are chosen because
they are the main consideration of the ultimate compen-
sation providers, that is, the shareholders of the rm, in
evaluating and approving compensation proposals. How-
ever, because the design of stock option plans contains
two main elements, that is, (i) the dilution of existing
equity holders and (ii) the price at which stock may be
purchased, it is straightforward to assume that these
attributes are simultaneously determined. A trade-off
exists between dilution and the associated premium.
Therefore, to allow for the possibility that the scope and
premium of stock option plans are determined simulta-
neously, we perform a simultaneous equation analysis
using three-stage least squares (3SLS), with the scope and
premium as endogenous variables. In particular, Eqs. (1)
and (2) are estimated concurrently. In this analysis, we
focus on the specications of the variables that are
expected to be the most critical elements, specically,
the stock option overhang and the (weighted) average
premium of the stock option plan.
Furthermore, we extend the analysis of the dynamics
of the stock option premium by estimating two categories
14
Since manager compensation consists of a combination of a xed
fee and a variable part (bonus plus executive options), especially the
option part might be considered as compensation for extraordinary
manager skill. Goetzmann, Ingersoll, and Ross (2003) characterize high-
water mark contracts as pure bets on manager skill, and it limits the
value of performance fees. Their analysis of the relative benets of
different fee structures in hedge fund industries suggests that high
variance strategies lend themselves to high-water mark contracting. In
our case, the analogy would be that especially in high volatility
industries, a high-water mark type of compensation contract might be
optimal.
of discrete decision models. Yermack (1997) argues that if
rms really do give extensive thought to the optimal
exercise prices of executive options, one should expect a
nontrivial fraction of options to be awarded out-of-the-
money. However, the literature shows that these pre-
mium options are extremely rare in the US, which
remains a puzzle. In Finland, the data show that a
nontrivial amount of stock options are awarded with
out-of-the-money exercise prices. The Finnish setting
thus offers an interesting laboratory to test the factors
associated with granting premium (out-of-the-money)
stock options. As a result, in the rst category, we
construct dependent variables that take the value of one
if (i) the rst tranche premium is positive and zero
otherwise, and (ii) the weighted average premium is
positive and zero otherwise. Furthermore, to provide
additional insights into the factors that drive awards of
premium stock options, we include an additional restric-
tion in the construction of the dependent variables. In the
second category, we construct dependent variables that
take the value of one if (i) the rst tranche premium is
positive and the stock option plan is targeted solely to the
top management of the rm and zero otherwise, and (ii)
the weighted average premium is positive for stock
options targeted to top management and zero otherwise.
Finally, we also test for the relation between the length
of the vesting period and the size of the premium to test
for whether in line with optimal contracting the
shareholders grant more out-of-the money options when
their vesting period is further in the future. This view is
contradictory to that of Bebchuk, Fried, and Walker (2002).
4. Sample characteristics
This section rst gives details of the Finnish market in
terms of corporate governance, business culture/share-
holder orientation, and executive compensation. Next, the
data sources used in the paper are described. Sample
selection issues are also discussed here. Finally, a descrip-
tive analysis of the data is presented.
4.1. Data sources and sample selection
In Finland, executive stock option plans (ESOPs) for
listed companies have to be approved by a shareholders
meeting, usually the Annual General Meeting (AGM). The
AGM typically approves both the scope of the plan
(number of options) and the conditions for setting the
exercise prices and vesting periods for the different
tranches of the program. This is unique; in other countries
decisions concerning contract details are typically left to
the board/compensation committee (see, e.g., Rosser and
Canil, 2004). In Finland, after the AGM approves of the
plan, the company board then decides, with the help of its
compensation committee (which consists of some inde-
pendent board members and excludes the CEO), the
structure of the CEOs compensation package and partici-
pates in setting the general guidelines for the rms
compensation policies. Alexander Corporate Finance Oy
has been a dominant consultant of companies in Finland
regarding ESOPs and thus has a unique data set of all of
the plans and their specic tranches. They also have
complemented their data set with information on the
few programs of listed rms that were not consulted by
Alexander Corporate Finance Oy. Our data set, which was
received from Alexander Corporate Finance Oy, thus
includes all stock option plans for all Finnish-listed rms
during the time period of 19872001. The data contain
information regarding the introduction date of stock
option plans, the target group, vesting periods and con-
tract maturities, the exercise prices, the number of shares
obtainable upon exercise of stock options, and whether
the stock options are subject to dividend protection and/
or performance-vesting/indexing.
Firm accounting data are obtained from the Research
Institute of the Finnish Economy (ETLA). Firm ownership
data are obtained from P orssiyhti ot manuals and the
Finnish Central Securities Depository (FCSD). Stock
returns, interest rates, and data regarding the number of
outstanding shares are obtained from the Hanken School
of Economics (Hanken). Stock prices and dividends are
collected from DataStream and Hankens database.
The raw sample covers the time period of 19872001
and consists of all public stock option plans in Finland
since the rst plan was introduced in 1987 (287 stock
option plans). The current study is restricted to rms that
trade on the main list of the Helsinki Stock Exchange
(HEX), with banks and insurance companies excluded from
the analysis. The initial sample is created by identifying
stock option plans for rms that have appeared on the
main list of HEX during the investigation period (the
presence of a trading code in Hankens database). Firms
may have launched several stock option plans during the
same year, and in these cases, we include only the rst
stock option plan that was launched during the year in the
analysis. To calculate the total Black-Scholes option value
of the stock option plans, we require a minimum of 60
daily stock returns prior to the grant date, which excludes
47 stock option plans fromthe analysis. The excluded plans
are those that were launched while the rm was listed on
other than the main list of HEX or that were launched in
direct connection with the rms initial public offering
(IPO) to the main list of HEX. Furthermore, to calculate
explanatory risk variables, we require a minimum of 60
daily stock returns prior to the end of the last accounting
period preceding the launch of the stock option plans,
which excludes an additional 13 stock option plans from
the analysis. This procedure results in a nal sample of 141
stock option plans for the period of 19872001, which
represents approximately 49% of the total number of stock
option plans introduced in Finland. Table 1 presents the
sample selection procedure used in the study.
Table 2 presents sample characteristics, divided into
Panel A (plan characteristics), Panel B (timing of stock option
plans), and Panel C (industry distribution). Of the 141 stock
option plans included in the sample, 106 (75%) were targeted
to the top management of the rm, whereas 35 (25%) stock
option plans were also targeted to non-executive employees
(broad-based stock option plans). Dividend protection is
included in 71 (50%) of the 141 stock option plans in the
study. Finally, of the sample stock option plans, 12 (9%)
involved performance-vesting/indexing, whereas 129 (91%)
stock option plans were traditional plans. In Finland, perfor-
mance-vesting typically implies that the exercise of stock
options is conditional on whether some accounting-based
measure of rm performance is fullled (e.g., net prot).
Indexing of stock options refers to the policy that the
exercise price of stock options is determined relative to, for
example, a peer-group benchmark. Table 2 shows the timing
of the stock option plans in the sample (introductions and
follow-up plans). As can be seen from Table 2, the majority
of stock option plans were granted in the latter half of the
1990s.
15
Finally, Table 2 also presents the distribution of stock
option plans across industries. The industry classications
are obtained from the accounting data of ETLA. The
industry is dened by ETLA as the area where at least
60% of annual sales are generated; otherwise, the rm is
classied as multi-business (diversied). Of the stock
option plans in the sample, 30 (21%) were granted by
diversied rms. Additionally, we divided the rms in the
focused category into mature and growth industries. The
statistics in Table 2 reveal that 72 (51%) of the stock
options plans in the sample were granted by rms in
mature industries, whereas 39 (28%) stock options were
launched by rms in growth industries. The median value
of Tobins Q for diversied rms, mature rms, and
growth rms is 1.113, 1.099, and 2.678, respectively.
4.2. Descriptive statistics
Table 3 provides a descriptive analysis of the data. The
statistics for the scope of stock option plans indicate that
the average value for the stock option overhang (dilution)
is 3.2% and that the average value of the total Black-Scholes
value divided by the market value of equity is 1.4%. As
mentioned earlier, the stock option overhang measures the
fraction of equity ownership that would result from the
exercise of all stock options. Furthermore, the values for
the stock option premium(out-of-the-moneyness) indicate
that the stock option plans in the sample have, on average,
been granted out-of-the-money. The average value for the
rst tranche premium is 10.4% and the weighted average
premiumis 11.2%, on average. The rst tranche premiumis
positive (out-of-the-money) in 69% of the sampled stock
option plans, zero (at-the-money) in 11% of the sampled
stock option plans, and negative (in-the-money) in 20% of
the sampled stock option plans. The corresponding gures
for the weighted average premium are 70% (out-of-the-
money), 11% (at-the-money), and 19% (in-the-money).
5. Empirical results
This section presents and discusses the empirical results.
First, we report and discuss the results regarding the
determinants of the scope, and the exercise price of stock
options, in Sections 5.1 and 5.2, respectively. We then report
results from further specication tests in Section 5.3.
5.1. Determinants of the scope of stock option plans
Table 4 presents regression results regarding the rela-
tions between rm attributes and the scope of stock
option plans. Panel A presents results from the regres-
sions where the scope of stock option plans is measured
as Stock option overhang, that is, the fraction of equity
obtained upon exercise of all granted stock options. Panel
A also reports results from four different specications:
specications [I] and [II] employ a measure of total rm
risk and specications [III] and [IV] decompose total rm
risk into systematic and unsystematic components. Spe-
cications [I] and [III] include an indicator variable for
rm focus, whereas specications [II] and [IV] decompose
this indicator variable into separate variables for mature
and growth industries.
16
Table 1
Sample selection.
The raw sample covers the time period 19872001 and consists of all
stock option plans introduced in Finland (287 stock option plans). The
current study is restricted to rms trading on the main list of the
Helsinki Stock Exchange (HEX), with banks and insurance companies
excluded from the analysis. Firms may have launched several stock
option plans during the same year, and in these cases, we include only
the initially launched stock option plan during the year in the analysis.
The sample selection procedure results in a nal sample of 141 stock
option plans during the years 19872001, which represents approxi-
mately 49% of the total number of stock option plans introduced in
Finland.
Sample selection
Total number of stock option plans in Finland (19872001) 287
Firms appearing on HEX main list during the sample period
(SSEBA-HEX code available)
226
Firms that have not appeared on HEX main list during the
sample period (I- or NM-lists: no SSEBA-HEX code available)
61
Total 287
Initial sample 1 226
Banks or insurance companies (HEX main list) 17
Multiple grants during year (HEX main list) 8
Lack of stock return history (a minimum of 60 daily
observations from grant date backwards) for Black and
Scholes option valuation
47
Firms may have launched stock option plans during listing on
other lists than HEX main list (years before listing on HEX
main list) or directly in connection with IPO to HEX main list
Lack of stock return history (a minimum of 60 daily
observations backwards from the last accounting period end
before grant date) for explanatory risk variables
13
Final sample (number of stock option plans) 141
Percentage of all stock option plans in Finland (19872001) 49%
15
Anecdotal evidence suggests that this phenomenon is driven by
factors such as the abolishment of restricting foreign ownership in
Finnish publicly traded rms in 1993, joining the European Monetary
Union (EMU), accompanied by the introduction of the euro in 1999, and
the favorable development of stock market valuations in the late 1990s.
16
Instead of the indicator variables for mature and growth indus-
tries, we re-estimated the models using dummies for seven industries.
The results were unaffected by this change. Also, the latter models were
re-estimated using industry dummies, without any signicant effects on
results.
The results reported in Panel A reveal that the variables
that measure the ownership structure of the rm generally
do not seem to be associated with the scope (stock option
overhang) of stock option plans. The coefcients for Tobins
Q are negative and signicant at the 1% level in all four
specications. As mentioned earlier, the predicted relation
between Tobins Q and the scope of stock option plans is
ambiguous and depends on whether one interprets Tobins
Q strictly as a proxy for the degree of growth opportunities
or for rm performance. If it proxies rm performance, the
result is consistent with the view that poorly performing
rms introduce larger stock option plans to enforce incen-
tive alignment (i.e., to motivate managers and employees
to maximize shareholder value).
The coefcient for Capital-to-sales is negative and system-
atically highly signicant at the 1% level. Furthermore, the
coefcients for Total risk are positive and signicant at the 5%
level. These results are in line with predictions from optimal
contracting theory, that is, consistent with the argument that
rms with a high degree of monitoring difculty/costs (a low
ratio of capital-to-sales and high variation in expected cash
ows) use greater equity incentives. Finally, the coefcients
for Broad-based plan and Dividend-protected plan are positive
and signicant at the 1% level in all specications. Although
several coefcients in Panel A exhibit statistical signicance,
it is noteworthy that the economic signicance of several
variables tends to be small.
Panels B and C display the results of the regression where
the scope of the stock option plan is measured as the ratio of
the total Black-Scholes value (Panel B) or the total Adjusted
Black-Scholes value (Panel C) to the market value of the
rms equity. Results for two different specications are
Table 2
Sample characteristics.
The sample covers the time period 19872001 and consists of stock option plans launched by rms traded on the main list of the Helsinki Stock
Exchange (HEX). Panel A. [1] displays the distribution of the target group of stock option plans. Stock option plans are dened as targeted top
management if stock options are targeted solely to the top management of the rm. If stock options are also targeted to non-executive employees, the
stock option plan is dened as being broad-based. Panel A. [2] presents the distribution of dividend protection in connection with stock option plans. In
dividend-protected stock option plans, exercise prices are adjusted (reduced) on the ex-dividend date for the amount of dividend payments per share.
Panel A. [3] shows the distribution of performance-vested/indexed stock option plans in the sample. A stock option plan is dened as performance-
vested/indexed if the exercise of stock options is conditional on whether some accounting-based measure of rm performance is fullled (e.g., net prot),
and/or whether the exercise price is determined relative to, e.g., a peer-group benchmark.
Panel A: Plan characteristics
A. [1] Target group A. [2] Dividend protection A. [3] Performance-vesting/indexing
Top management plans 106 (75%) Included 71 (50%) Included 12 (9%)
Broad-based plans 35 (25%) Not included 70 (50%) Not included 129 (91%)
Total 141 Total 141 Total 141
Panel B: Timing of stock option plans
B. [1] Introductions B. [2] Follow-up contracts B. [3] New stock option plans
1987 1 1987 0 1987 1
1988 2 1988 0 1988 2
1989 3 1989 2 1989 5
1990 1 1990 1 1990 2
1991 0 1991 2 1991 2
1992 1 1992 0 1992 1
1993 4 1993 1 1993 5
1994 12 1994 2 1994 14
1995 2 1995 2 1995 4
1996 2 1996 4 1996 6
1997 5 1997 11 1997 16
1998 7 1998 15 1998 22
1999 3 1999 11 1999 14
2000 6 2000 21 2000 27
2001 1 2001 19 2001 20
Total 50 91 141
Panel C: Industry distribution
Chemical and plastic [M] 9 (6%) Foods [M] 10 (7%) Retail [M] 3 (2%)
Construction [M] 5 (4%) Forest [M] 8 (6%) Services [M] 1 (1%)
Diversied [D] 30 (21%) Information technology [G] 13 (9%) Telecommunications [G] 3 (2%)
Electricity and electronical equipment [G] 7 (5%) Investment [G] 5 (4%) Textile [M] 3 (2%)
Electronical equipment [G] 7 (5%) Media [G] 4 (3%) Transportation [M] 6 (4%)
Energy [M] 1 (1%) Metal [M] 20 (14%) Wholesale [M] 6 (4%)
Diversied [D] 30 21%
Mature industry [M] 72 51%
Growth industry [G] 39 28%
Total 141 100%
reported in both panels. In the rst specication [I], a rm-
focus indicator variable is included, whereas the second
specication [II] includes separate variables for mature and
growth industries. The specications in Panels B and C do not
include variables that measure rm risk because dependent
variables (the BS value or the Adjusted BS value, which are
both related to the MV of equity) incorporate the historical
standard deviation of stock returns in their measurement.
In line with the results in Panel A, Tobins Q is also
negative and signicant in Panel C, but not in Panel B.
Capital-to-sales and Broad-based plan also obtain the
expected signs in Panels B and C and are again highly
signicant. In Panel B, the coefcient of Dividend-protected
plan is also positive and signicant as before, now at the
5% level, but the coefcient is not signicant in specica-
tion [III] in Panel C.
The results in Panels B and C differ from those in Panel A
in some respects. Long-term debt-to-assets is now signi-
cant at levels ranging from 1% to 10% in the different
specications. This contradicts the view that debt sub-
stitutes for equity incentives (Jensen, 1986, 1993) and
with John and Johns (1993) prediction that highly levered
rms will provide less equity incentives to motivate
optimal managerial risk choices. Both Non-state ownership
control as well as State ownership are signicant with
expected signs in Panel C, which indicates that stronger
ownership control is associated with smaller option
grants. Also in Panel C, Cash ow-to-assets is signicant
Table 3
Descriptive statistics.
The sample covers the time period 19872001 and consists of rms traded on the main list of the Helsinki Stock Exchange (HEX). The table presents
summary statistics for the full sample. Stock option overhang is calculated as the total number of shares exercisable induced by the stock option plan
divided by the sum of the total number of shares exercisable and the number of shares outstanding at the grant date. Stock options are valued using the
Black-Scholes (1973) methodology adjusted for dividend payments as in Merton (1973). Stock options specically protected against dividend payments
are valued using standard Black-Scholes methodology. The stock option premium (out-of-the-moneyness) is dened as [(XS)/S], where X corresponds
to the exercise price of the option and where S is the stock price at the grant date. The rst-tranche stock option premium is calculated as the out-of-the-
moneyness of the stock options belonging to the rst tranche in the stock option plan. The weighted average stock option premium utilizes information
of the characteristics of the total stock option plan. The weights used in the calculation correspond to the ratio of the number of shares obtainable upon
exercise in each individual stock option tranche divided by the total number of shares obtainable upon exercise of all stock options. Data on foreign
ownership are only available from the Finnish Central Securities Depositary (FCSD) from October 1993 onwards, and due to this, data on foreign
ownership are missing in 24 observations. Complete data on prior stock returns are missing in four observations. All risk measures are multiplied by 100.
See Appendix A for a denition of variables.
Variable Mean Median Std. dev. 1st Quart. 3rd Quart. Minimum Maximum
Number of observations [141]
Stock option plan characteristics
Stock option overhang [shares exercisable/(shares
exercisableshares outstanding)]
0.032 0.026 0.023 0.017 0.045 0.001 0.108
Total Black and Scholes value of option plan [h] 45,389,440 3,024,407 278,188,950 1,346,701 8,726,311 3,315 2,958 10
6
Total Black and Scholes value of option plan to
market value of equity
0.014 0.009 0.014 0.004 0.019 0.000 0.081
Adjusted total Black and Scholes value of option
plan to market value of equity
0.007 0.005 0.006 0.002 0.009 0.000 0.031
First tranche premium [(XS)/S] 0.104 0.064 0.172 0.000 0.180 0.251 0.869
Weighted average premium [(XS)/S] 0.112 0.078 0.178 0.000 0.189 0.251 0.869
Ownership variables
CEO ownership 0.011 0.000 0.053 0.000 0.000 0.000 0.440
Non-state ownership control 0.316 0.286 0.172 0.172 0.433 0.018 0.799
Institutional ownership [1/0] 0.887
State ownership [1/0] 0.142
Foreign ownership [117 observations] 0.272 0.207 0.235 0.069 0.448 0.001 0.960
Firm characteristics
Total assets [h000] 1 725 771 586 471 3 526 308 135 559 1 396 000 20 014 21 322 865
Tobins Q 2.335 1.202 3.679 0.997 1.973 0.715 30.928
Investment-to-capital 0.274 0.214 0.201 0.128 0.369 0.019 1.011
Long-term debt-to-assets 0.224 0.220 0.144 0.099 0.332 0.001 0.647
Wages per employee 0.207 0.204 0.069 0.161 0.248 0.053 0.446
Zero-dividends [1/0] 0.142
Cash ow-to-assets 0.144 0.126 0.085 0.095 0.170 0.059 0.462
Free cash ow-to-assets 0.002 0.017 0.127 0.030 0.067 0.804 0.300
Prior stock return [137 observations] 0.066 0.057 0.304 0.120 0.266 0.818 1.127
Capital-to-sales 0.877 0.433 2.017 0.252 0.682 0.057 13.577
Firm focus [1/0] 0.787
Mature industry [1/0] 0.511
Growth industry [1/0] 0.277
Total risk 0.077 0.056 0.069 0.038 0.094 0.013 0.479
Systematic risk 0.014 0.006 0.025 0.002 0.015 0.000 0.157
Unsystematic risk 0.063 0.048 0.051 0.031 0.078 0.012 0.331
Number of rms 75
Table 4
Determinants of the scope of stock option plans.
The stock option overhang in Panel A is measured as the fraction of equity obtained upon exercise of all granted stock options, i.e., as the ratio of the number
of shares exercisable to the sum of shares exercisable and the number of outstanding shares at the date of the grant. In Panel B the scope of the stock option
plan is calculated as the total Black-Scholes value divided by the market value of equity at the grant date. Panel C uses Black-Scholes values adjusted following
the method suggested by Meulbroek (2001) for a fully undiversied manager and divided by the market value of equity at the grant date. The independent
variables are measured at the end of the previous year. A full set of year dummies is included in all specications (not reported). All risk measures are
multiplied by 100. See Appendix A for variable denitions. Estimation is conducted utilizing OLS. The t-statistics (reported beneath each regression coefcient)
are calculated using robust standard errors. The 1%, 5%, and 10% signicance levels are denoted with
nnn
,
nn
, and
n
, respectively. Final sample: Stock option plans
introduced by rms trading on the main list of the Helsinki Stock Exchange (HEX) during the years 19872001 (see Table 1).
Independent variables Dependent variable
Panel A: Stock option overhang
[I] [II] [III] [IV]
CEO ownership 0.025
n
0.025
n
0.022 0.022
(1.73) (1.72) (1.47) (1.46)
Non-state ownership control 0.003 0.003 0.003 0.003
(0.27) (0.27) (0.26) (0.26)
Institutional ownership [1/0] 0.002 0.003 0.003 0.003
(0.36) (0.35) (0.42) (0.42)
State ownership [1/0] 0.007 0.007 0.007 0.007
(1.35) (1.36) (1.37) (1.38)
Firm size 0.002
n
0.002 0.002
n
0.002
n
(1.66) (1.63) (1.74) (1.73)
Tobins Q 0.003
nnn
0.003
nnn
0.003
nnn
0.003
nnn
(3.82) (3.76) (3.72) (3.77)
Investment-to-capital 0.017 0.017 0.016 0.016
(1.36) (1.37) (1.26) (1.27)
Long-term debt-to-assets 0.016 0.016 0.018 0.018
(1.17) (1.11) (1.25) (1.21)
Cash ow-to-assets 0.030 0.029 0.026 0.026
(1.08) (1.03) (0.95) (0.91)
Capital-to-sales 0.003
nnn
0.003
nnn
0.003
nnn
0.003
nnn
(3.68) (3.37) (3.64) (3.35)
Firm focus [1/0] 0.003 0.003
(0.67) (0.65)
Mature industry [1/0] 0.003 0.003
(0.68) (0.66)
Growth industry [1/0] 0.002 0.002
(0.33) (0.30)
Total risk 0.081
nn
0.081
nn
(2.23) (2.18)
Systematic risk 0.155 0.155
(1.21) (1.21)
Unsystematic risk 0.054 0.053
(1.01) (1.01)
Prior plan in effect [1/0] 0.002 0.002 0.001 0.001
(0.47) (0.46) (0.39) (0.39)
Broad-based plan [1/0] 0.025
nnn
0.024
nnn
0.025
nnn
0.025
nnn
(5.37) (5.07) (5.45) (5.12)
Dividend-protected plan [1/0] 0.011
nnn
0.011
nnn
0.012
nnn
0.012
nnn
(2.86) (2.88) (2.87) (2.90)
PV/indexed plan [1/0] 0.0004 0.0004 0.0004 0.0004
(0.09) (0.09) (0.08) (0.08)
Year dummies Yes Yes Yes Yes
Adjusted R-squared 0.488 0.483 0.486 0.481
Number of observations 141 141 141 141
Dependent variable
Panel B: BS value to MV of equity Panel C: Adjusted BS value to MV of equity
[I] [II] [III] [IV]
CEO ownership 0.004 0.006 0.001 0.003
(0.44) (0.59) (0.24) (0.51)
Non-state ownership control 0.001 0.001 0.005
n
0.005
n
(0.14) (0.11) (1.77) (1.84)
Institutional ownership [1/0] 0.004 0.003 0.0001 0.0011
(0.83) (0.63) (0.05) (0.66)
with a positive sign. This contradicts the hypothesis that
Cash ow-to-assets proxies protability, and higher prot-
ability would be associated with less need for strong
incentives if the expected marginal benets of equity-
based compensation are decreasing in rm performance.
If Cash ow-to-assets instead proxies for agency costs,
greater pay-performance sensitivity might be motivated
in high cash ow rms to reduce agency costs.
In summary, as determinants for the scope of the
option plan, we get the strongest results for two variables
derived from the optimal contracting theory: the Capital-
to-sales ratio as a proxy for monitoring complexity (sig-
nicantly negative in all specications), which suggests
that options are used more in more complex rms, and
Tobins Q (systematically negative, and signicant in the
stock overhang specications), which may, in line with,
e.g., Morck, Shleifer, and Vishny (1988), be interpreted as
a proxy for rm performance, suggesting a greater need to
incentivize managers in poorly performing rms where
the expected marginal benets may be larger.
5.2. Factors driving the exercise price of stock options
Table 5 reports regression results for determinants of
the stock option premium (the out-of-the-moneyness). In
Panel A of Table 5, the dependent variable is the First
tranche premium, whereas in Panel B, we use the Weighted
average premium. The table reports four different speci-
cations in each panel: either including a measure of Total
risk (specications [I] and [II]), or with rm risk decom-
posed into Systematic risk and Unsystematic risk (specica-
tions [III] and [IV]). Furthermore, the specications either
utilize an indicator variable for Firm focus (specications
[I] and [III]) or decompose this measure into two separate
variables for Mature industries and Growth industries
(specications [II] and [IV]). Data for Prior stock return
are missing in four observations. We apply the same
method that is used by Himmelberg, Hubbard, and Palia
(1999) and Jin (2002), among others; that is, we set
missing observations of prior stock returns to zero and
include a separate indicator variable that takes the value
of one for missing observations and zero otherwise.
The results in Table 5 reveal a similar picture as in
Table 4; that is: that ownership structure does not seem
to be related to the design of stock option plans. Further,
as in the case of scope, very few potential other determi-
nants obtain signicance. One exception is Prior stock
return, which is inversely related to the stock option
premium and signicant at the 5% or 10% level through-
out the specications. This result supports both the
Table 4 (continued )
Dependent variable
Panel B: BS value to MV of equity Panel C: Adjusted BS value to MV of equity
[I] [II] [III] [IV]
State ownership [1/0] 0.004 0.004 0.003
n
0.003
n
(1.56) (1.53) (1.79) (1.8)
Firm size 0.001 0.001 0.0004 0.0005
(0.96) (0.91) (1.34) (1.49)
Tobins Q 0.001 0.001 0.0004
n
0.0005
nnn
(1.51) (1.63) (1.8) (2.73)
Investment-to-capital 0.017
n
0.014 0.007 0.004
(1.67) (1.52) (1.64) (0.98)
Long-term debt-to-assets 0.014
n
0.016
nn
0.001
nn
0.012
nnn
(1.82) (2.13) (2.07) (2.61)
Cash ow-to-assets 0.003 0.005 0.021
nnn
0.018
nn
(0.21) (0.39) (2.95) (2.43)
nnn
0.003
nnn
0.001
nn
0.001
nnn
(5.11) (4.81) (2.38) (3.41)
Firm focus [1/0] 0.003 0.001
(1.27) (0.82)
Mature industry [1/0] 0.003 0.002
(1.60) (1.56)
Growth industry [1/0] 0.001 0.003
(0.36) (1.6)
Prior plan in effect [1/0] 0.002 0.002 0.000 0.0000
(1.27) (1.23) (0.01) (0.03)
nnn
0.013
nnn
0.007
nnn
0.006
nnn
(5.27) (4.57) (5.09) (4.56)
nn
0.005
nn
0.002 0.002
n
(2.12) (2.26) (1.46) (1.81)
PV/indexed plan [1/0] 0.001 0.001 0.002 0.002
(0.55) (0.55) (0.75) (0.81)
Year dummies Yes Yes Yes Yes
Adjusted R-squared 0.462 0.468 0.458 0.514
Number of observations 141 141 141 141
optimal contracting theory (high-water mark contracting)
and the managerial power hypothesis, because both of
these theories predict a negative sign. In terms of eco-
nomic signicance, the results in Panel B suggest that a
unit increase in the prior (six-month) stock return
translates into a stock option premium that is approxi-
mately 0.2 percentage points lower.
We perform additional tests to distinguish between the
two alternative explanations by estimating the models in
Table 5 only for the observations with a positive prior
Table 5
Determinants of the stock option premium.
The stock option premium (out-of-the-moneyness) is dened as [(XS)/S], where X corresponds to the exercise price of the option and where S is the
stock price at the grant date. In Panel A, the rst tranche premium is calculated as the out-of-the-moneyness of the stock options belonging to the rst
tranche. In Panel B, the weighted average stock option premium is used. The weights used in the calculation correspond to the ratio of the number of
shares obtainable upon exercise in each individual stock option tranche divided by the total number of shares obtainable upon exercise of all stock
options. The independent variables are measured at the end of the previous year. A full set of year dummies is included in all specications (not
reported). All risk measures are multiplied by 100. Data for prior stock returns are missing in four observations. See Appendix A for variable denitions.
Estimation is conducted utilizing OLS. The t-statistics (reported beneath each regression coefcient) are calculated using robust standard errors. The 1%,
5%, and 10% signicance levels are denoted with
nnn
,
nn
, and
n
, respectively. Final sample: Stock option plans introduced by rms trading on the main list
of the Helsinki Stock Exchange (HEX) during the years 19872001 (see Table 1).
Panel A: First tranche premium Panel B: Weighted average premium
[I] [II] [III] [IV] [I] [II] [III] [IV]
CEO ownership 0.008 0.006 0.028 0.015 0.007 0.011 0.010 0.014
(0.03) (0.02) (0.10) (0.05) (0.03) (0.04) (0.04) (0.05)
Non-state ownership control 0.052 0.051 0.054 0.053 0.046 0.046 0.046 0.045
(0.47) (0.46) (0.49) (0.48) (0.41) (0.40) (0.40) (0.40)
Institutional ownership [1/0] 0.018 0.011 0.014 0.007 0.030 0.028 0.030 0.028
(0.38) (0.24) (0.29) (0.15) (0.64) (0.57) (0.66) (0.59)
State ownership [1/0] 0.018 0.020 0.018 0.020 0.028 0.029 0.028 0.029
(0.35) (0.38) (0.36) (0.39) (0.52) (0.53) (0.52) (0.53)
Firm size 0.002 0.002 0.001 0.001 0.004 0.004 0.004 0.004
(0.20) (0.23) (0.09) (0.07) (0.35) (0.34) (0.26) (0.25)
Tobins Q 0.005 0.005 0.004 0.004 0.005 0.005 0.005 0.005
(0.80) (0.69) (0.54) (0.44) (0.77) (0.71) (0.70) (0.63)
Investment-to-capital 0.114 0.134 0.120 0.140 0.063 0.070 0.062 0.069
(1.01) (1.13) (1.02) (1.13) (0.63) (0.65) (0.59) (0.61)
Long-term debt-to-assets 0.016 0.0003 0.009 0.007 0.010 0.005 0.011 0.006
(0.11) (0.01) (0.06) (0.05) (0.07) (0.03) (0.08) (0.04)
Cash ow-to-assets 0.374 0.350 0.344 0.319 0.363
n
0.355
n
0.367
n
0.359
(1.48) (1.38) (1.32) (1.21) (1.76) (1.71) (1.73) (1.65)
Prior stock return 0.144
nn
0.135
n
0.154
nn
0.146
n
0.182
nn
0.179
nn
0.181
nn
0.178
nn
(2.06) (1.80) (2.11) (1.90) (2.26) (2.08) (2.24) (2.09)
Prior stock return missing [1/0] 0.104 0.101 0.101 0.098 0.070 0.069 0.070 0.069
(1.57) (1.39) (1.58) (1.40) (0.97) (0.91) (0.96) (0.90)
Capital-to-sales 0.015 0.018 0.015 0.018 0.018
n
0.019 0.018
n
0.019
(1.37) (1.52) (1.42) (1.57) (1.71) (1.66) (1.69) (1.64)
Firm focus [1/0] 0.009 0.009 0.020 0.020
(0.27) (0.26) (0.54) (0.53)
Mature industry [1/0] 0.015 0.015 0.022 0.022
(0.42) (0.41) (0.58) (0.58)
Growth industry [1/0] 0.024 0.025 0.009 0.009
(0.46) (0.46) (0.15) (0.15)
Total risk 0.481 0.508 0.599
n
0.608
n
(1.28) (1.38) (1.68) (1.73)
Systematic risk 0.021 0.044 0.661 0.668
(0.02) (0.04) (0.55) (0.57)
Unsystematic risk 0.649 0.678 0.576 0.586
(1.23) (1.30) (1.10) (1.11)
Prior plan in effect [1/0] 0.004 0.005 0.002 0.003 0.001 0.001 0.001 0.001
(0.13) (0.14) (0.07) (0.08) (0.02) (0.02) (0.02) (0.03)
Broad-based plan [1/0] 0.027 0.035 0.027 0.034 0.037 0.039 0.037 0.039
(0.61) (0.77) (0.59) (0.75) (0.79) (0.86) (0.79) (0.86)
Dividend-protected plan [1/0] 0.051 0.050 0.046 0.045 0.074
nn
0.074
nn
0.075
nn
0.075
nn
(1.28) (1.27) (1.21) (1.21) (2.12) (2.12) (2.28) (2.27)
PV/indexed plan [1/0] 0.073 0.073 0.073 0.074 0.082 0.082 0.082 0.082
(1.52) (1.54) (1.53) (1.54) (1.55) (1.55) (1.54) (1.54)
Year dummies Yes Yes Yes Yes Yes Yes Yes Yes
Adjusted R-squared 0.218 0.216 0.213 0.211 0.262 0.256 0.255 0.249
Number of observations 141 141 141 141 141 141 141 141
return. If high-water mark contracting is the reason for the
relation reported above between option premium and the
prior stock return, such a relation should not be present in
the subsample of rms with a positive stock-price history.
Using two alternative denitions for a positive prior return
(one year vs. six months and 90 vs. 81 observations with
positive returns), we still obtain negative signs for Prior stock
return in the different specications of Table 5, but the
variable is no longer signicant (t-values around 1.0).
17
In
summary, our test does not rule out the possibility that the
observed relation is a result of high-water mark contracting.
The coefcients for the Cash ow-to-assets ratio are
systematically negative and statistically signicant in
Panel B (at the 10% level in specications [I] through
[III]). Although it is not very robust, this result suggests
that rms with higher performance set lower exercise
prices for stock options, which provides additional sup-
port for the managerial power hypothesis.
Furthermore, the coefcients for Total risk are positive
and signicant at the 10% level in specications [I] and [II]
of Panel B. This observation is in line with the positive
sign under one of the two optimal contracting alternatives
(high-water mark contracting), and in contrast to the
other, that is, the view that because rm risk proxies for
monitoring difculty, riskier rms would introduce stock
options with lower premiums. Finally, the coefcients for
Dividend-protected plan are positive and signicant
throughout the specications in Panel B. These results
are consistent with the expectations in that, all other
things being equal, dividend protection increases the
expected value of the stock option and thus reduces the
negative effect of the higher premium.
In summary, we nd the Prior stock return to be the
strongest determinant of the stock option premium, and
we also get some support for a performance variable, Cash
ow-to-assets. While the rst result may be in line with
optimal contracting (high-water mark contracts), a view
also supported by the positive sign of Total risk, the latter
result also lends some support to the managerial power
hypothesis.
5.3. Additional tests regarding scope and exercise prices
Estimation results from a simultaneous equation ana-
lysis are reported in Table 6. Stock option overhang and
Weighted average premium are treated as endogenous
variables.
18
Supporting our expectations, the results
suggest that the Stock option premium is positively related
to the scope, and vice versa. However, the results are not
statistically signicant at conventional levels.
More importantly, the results regarding the scope of
stock option plans in Table 4 are supported by the results
in Table 6. Specically, the results suggest that the scope
is decreasing in Tobins Q and Capital-to-sales and is
increasing in Total risk. Additionally, the results reveal
that the scope is greater in plans that are targeted to a
broader base of employees
19
and that the scope tends to
be larger in plans that involve dividend protection.
In the case of the stock option premium, the results are
similar to the corresponding regression results in Table 5.
However, the coefcients for Cash ow-to-assets and Capital-
to-sales gain higher statistical signicance in the simulta-
neous equation analysis. Most importantly, the documented
negative impact of Prior stock return on option premium is
corroborated by the results in Table 6, where the coefcient
is negative and signicant at the 1% level. The coefcient for
Dividend-protected plan in the premium specication is
positive; however, in this case, it is not statistically sig-
nicant at conventional levels.
To further elaborate on the factors that drive exercise
prices, we estimate discrete (Probit) decision models that
predict the launch of premium (out-of-the-money) stock
option plans. The results of these analyses are reported in
Tables 7 and 8. The dependent variables are constructed on
the basis of the (continuous) dependent variables in Table 5,
that is, the First tranche premium and the Weighted average
premium. In Table 7, the dependent variable takes the value
of one if the plan is launched out-of-the-money and zero
otherwise. The dependent variable in Table 8 is constructed
with an additional restriction such that it takes the value of
one if the stock option plan is both granted out-of-the-
money and targeted at the top management of the rm and
zero otherwise. Panel A of Tables 7 and 8 reports the results
from the regression where the dependent variable is based
on the First tranche premium, whereas Panel B of Tables 7
and 8 reports corresponding results, where the dependent
variable is based on the Weighted average premium.
20
Inspection of the results regarding ownership structure
in Panels A and B of Table 7 supports the same conclusion as
in the previous analyses: ownership structure does not
seem to be related to the design of stock option plans.
In line with our expectations for the premium, the
coefcients for the Investment-to-capital ratio are negative
and statistically signicant throughout the specications.
17
If option premium is regressed on Prior stock return without other
control variables, the coefcients are signicantly negative in all sub-
samples and return denitions. When only the two rm variables
exhibiting the highest t-values, together with option plan characteris-
tics, are included, Prior stock return is also negative and signicant in all
but one case.
18
The explanatory variables correspond to the same variables as in
Eq. (1) and (2), with the exception that an indicator variable for broad-
based plans is excluded from the premium specication. This is done to
facilitate identication in the simultaneous equation system. This choice
was made on the basis of the previous results indicating systematic
differences in the scope of stock option plans depending on the target
group of plans. However, the results in Table 6 were also found to be
robust to different choices of explanatory variables.
19
We also estimated a model for the determinants of the likelihood
of launching broad-based option plans (using a Probit model). The
results (not reported here but available from the authors) suggest that
institutional ownership signicantly increases the likelihood of granting
broad-based stock option plans. We also nd that the likelihood of
broad-based plans is signicantly greater among rms with high growth
opportunities, a higher degree of human capital, and more cash ow
constraints (a lower free cash ow).
20
The regressions in Table 7 are estimated for a marginally reduced
sample (137 observations), as the dependent variable in four observa-
tions lacking data on prior stock return is equal to unity. The regressions
in Table 8 are estimated for the full sample (141 observations) utilizing
the procedure of Himmelberg, Hubbard, and Palia (1999) and Jin (2002),
among others, for dealing with missing data.
Choe (2003) shows, all other things being equal, that the
optimal exercise price of stock options is decreasing in the
riskiness of the rms desired investment policy. The
result appears consistent with this view if one assumes
that investment intensity is a proxy for the riskiness of
the rms investment policy. Furthermore, in support of
our expectations and prior ndings for the premium
(in Table 5), the results in Table 7 indicate that the
Table 6
Simultaneous equation analysis of scope and premium of stock option plans.
The table reports estimation results using the three-stage least squares (3SLS) method. The scope of the stock
option plans (stock option overhang) is measured as the fraction of equity obtained upon exercise of all granted
stock options, i.e., as the ratio of the number of shares exercisable to the sum of shares exercisable and the
number of outstanding shares at the date of the grant. Weighted average stock option premium (out-of-the-
moneyness) is dened as [(XS)/S], where X corresponds to the exercise price of the option and where S is the
stock price at the grant date. The weights used in the calculation correspond to the ratio of the number of shares
obtainable upon exercise in each individual stock option tranche divided by the total number of shares obtainable
upon exercise of all stock options. The independent variables are measured at the end of the previous year. A full
set of year dummies is included in all specications (not reported). All risk measures are multiplied by 100. Data
for prior stock returns are missing in four observations. We apply the method utilized by Himmelberg, Hubbard,
and Palia (1999) and Jin (2002), among others; i.e., we set missing observations of prior stock returns to zero and
include a separate indicator variable that takes the value of one for missing observations and zero otherwise.
The t-statistics (reported beneath each regression coefcient) are calculated using robust standard errors.
See Appendix A for variable denitions. The 1%, 5%, and 10% signicance levels are denoted with
nnn
,
nn
, and
n
,
respectively. Final sample: Stock option plans introduced by rms trading on the main list of the Helsinki Stock
Exchange (HEX) during the years 19872001 (see Table 1).
Stock option
overhang
Weighted average
premium
Weighted average premium 0.001
(0.03)
Stock option overhang 1.488
(1.01)
CEO ownership 0.025 0.040
(0.97) (0.16)
Non-state ownership control 0.003 0.051
(0.30) (0.57)
Institutional ownership [1/0] 0.002 0.027
(0.49) (0.59)
State ownership [1/0] 0.007 0.019
(1.29) (0.39)
Firm size 0.002
n
0.008
(1.72) (0.65)
Tobins Q 0.003
nnn
0.001
(4.90) (0.24)
n
0.038
(1.92) (0.43)
Long-term debt-to-assets 0.016 0.011
(1.45) (0.10)
Cash ow-to-assets 0.030 0.396
nn
(1.22) (2.02)
Prior stock return 0.186
nnn
(3.30)
Prior stock return missing [1/0] 0.053
(0.64)
nn
0.022
nnn
(2.50) (2.84)
Firm focus [1/0] 0.003 0.024
(0.70) (0.70)
Total risk 0.081
nn
0.472
n
(2.51) (1.66)
Prior plan in effect [1/0] 0.002 0.003
(0.48) (0.09)
nnn
(6.46)
nn
0.057
(2.30) (1.43)
PV/indexed plan [1/0] 0.0003 0.081
(0.06) (1.56)
Year dummies Yes Yes
Adjusted R-squared 0.482 0.215
Number of observations 141 141
likelihood of granting premium (out-of-the-money) stock
options is decreasing in rm protability (measured as
the ratio of Cash ow-to-assets) because the coefcients of
the Cash ow-to-assets ratio are negative and statistically
signicant in all specications.
Finally, the coefcients of Capital-to-sales are again
positive (as expected and obtained for the premium in
Table 5) and now statistically signicant in all specications
reported in Table 7. We have interpreted Capital-to-sales as a
proxy for monitoring complexity. Lower monitoring costs
(a higher Capital-to-sales) seem to be associated with a
higher likelihood of granting premium (out-of-the-money)
stock options, which is in line with our expectations because
on a share-for-share basis, the pay-for-performance
Table 7
Probit estimation results predicting the launch of premium stock option plans.
The stock option premium (out-of-the-moneyness, OTM) is dened as [(XS)/S], where X corresponds to the exercise price of the option and where S is
the stock price at the grant date. In Panel A, the dependent variable takes the value of one if the rst tranche premium is positive and zero otherwise. In
Panel B, the dependent variable takes the value of one if the weighted average premium is positive and zero otherwise. The independent variables are
measured at the end of the previous year. All risk measures are multiplied by 100. Final sample: Stock option plans introduced by rms trading on the
main list of the Helsinki Stock Exchange (HEX) during the years 19872001 (see Table 1). The dependent variable in four observations lacking data on
prior stock return is set equal to unity. The regressions are therefore estimated for the reduced sample (137 observations) including complete data on
prior stock returns. See Appendix A for variable denitions. The t-statistics (reported beneath each regression coefcient) are calculated using robust
standard errors. The 1%, 5%, and 10% signicance levels are denoted with
nnn
,
nn
, and
n
, respectively.
Panel A: First tranche premium OTM [1/0] Panel B: Weighted average premium OTM [1/0]
CEO ownership 0.924 0.997 1.432 1.501 1.123 1.210 1.753 1.841
(0.43) (0.46) (0.65) (0.68) (0.51) (0.55) (0.80) (0.84)
Non-state ownership control 0.942 0.952 0.840 0.854 1.100 1.115 0.993 1.012
(1.09) (1.11) (0.95) (0.97) (1.26) (1.29) (1.10) (1.13)
(0.21) (0.31) (0.02) (0.07) (0.13) (0.24) (0.17) (0.05)
State ownership [1/0] 0.170 0.183 0.190 0.203 0.271 0.287 0.297 0.315
(0.40) (0.43) (0.44) (0.47) (0.63) (0.67) (0.67) (0.72)
Firm size 0.032 0.032 0.114 0.114 0.046 0.047 0.145 0.146
(0.30) (0.31) (0.97) (0.97) (0.44) (0.44) (1.22) (1.22)
Tobins Q 0.027 0.029 0.008 0.007 0.030 0.032 0.011 0.009
(0.55) (0.59) (0.16) (0.12) (0.60) (0.65) (0.20) (0.17)
nn
1.364
n
1.627
nn
1.457
nn
1.673
nn
1.474
n
1.785
nnn
1.588
nn
(2.27) (1.84) (2.45) (2.00) (2.42) (1.95) (2.66) (2.15)
(0.40) (0.53) (0.35) (0.47) (0.19) (0.34) (0.12) (0.26)
Cash ow-to-assets 3.553
n
3.435
n
4.479
nn
4.363
nn
3.993
nn
3.861
nn
5.137
nnn
5.010
nnn
(1.87) (1.80) (2.41) (2.33) (2.08) (2.01) (2.79) (2.71)
Prior stock return 0.538 0.481 0.328 0.273 0.426 0.358 0.169 0.100
(1.12) (1.00) (0.65) (0.54) (0.87) (0.74) (0.33) (0.20)
nn
0.161
nn
0.163
nn
0.175
nn
0.147
nn
0.164
nnn
0.169
nn
0.182
nn
(2.58) (2.55) (2.39) (2.48) (2.59) (2.62) (2.33) (2.48)
Firm focus [1/0] 0.031 0.0002 0.045 0.085
(0.09) (0.01) (0.14) (0.25)
Mature industry [1/0] 0.007 0.037 0.090 0.130
(0.02) (0.11) (0.27) (0.38)
Growth industry [1/0] 0.237 0.202 0.194 0.154
(0.48) (0.40) (0.39) (0.31)
Total risk 0.494 0.228 0.088 0.230
(0.22) (0.10) (0.04) (0.10)
Systematic risk 10.967 11.198 13.469 13.812
(1.21) (1.23) (1.47) (1.50)
(1.16) (1.07) (1.20) (1.10)
(0.76) (0.74) (0.43) (0.41) (0.90) (0.87) (0.52) (0.49)
Broad-based plan [1/0] 0.063 0.119 0.098 0.155 0.014 0.082 0.057 0.127
(0.18) (0.33) (0.28) (0.42) (0.04) (0.22) (0.16) (0.35)
Dividend-protected plan [1/0] 0.367 0.353 0.419 0.407 0.355 0.339 0.420 0.406
(1.23) (1.18) (1.39) (1.35) (1.17) (1.11) (1.38) (1.33)
PV/indexed plan [1/0] 0.535 0.531 0.511 0.508 0.509 0.504 0.485 0.482
(1.13) (1.13) (1.04) (1.05) (1.08) (1.08) (0.98) (0.99)
Log likelihood 75.518 75.356 74.645 74.494 73.999 73.782 72.803 72.594
Pseudo R-squared 0.122 0.124 0.132 0.134 0.132 0.134 0.146 0.148
Dependent variable [at 1 in %] 0.679 0.679 0.679 0.679 0.686 0.686 0.686 0.686
sensitivity of out-of-the-money stock options is lower than
that of corresponding in-the-money or at-the-money stock
options (see, e.g., Lambert, Larcker, and Verrecchia, 1991).
Regression results that predict the launch of premium
stock options targeted at top management are reported in
Table 8. The dependent variable in Table 8 takes the value
of one if the stock option plan is launched out-of-the-
money and is targeted at the top management, and zero
otherwise. The results in Table 8 are generally similar to
the results reported in Table 7. First, the coefcients of the
Investment-to-capital are negative and statistically signi-
cant throughout the specications in Table 8 and thus
Table 8
Probit estimation results predicting the launch of premium stock option plans targeted to top management.
The stock option premium (out-of-the-moneyness) is dened as [(XS)/S], where X corresponds to the exercise price of the option and where S is the
stock price at the grant date. In Panel A, the dependent variable takes the value of one if the rst tranche premium is positive and the plan is targeted
solely to top management and zero otherwise. In Panel B, the dependent variable takes the value of one if the weighted average premium is positive and
the plan is targeted solely to top management and zero otherwise. The independent variables are measured at the end of the previous year. All risk
measures are multiplied by 100. Data for prior stock return are missing in four observations. See Appendix A for variable denitions. The t-statistics
(reported beneath each regression coefcient) are calculated using robust standard errors. The 1%, 5%, and 10% signicance levels are denoted with
nnn
,
nn
,
and
n
, respectively. Final sample: Stock option plans introduced by rms trading on the main list of the Helsinki Stock Exchange (HEX) during the years
19872001 (see Table 1).
Panel A: First tranche premium OTM in plans
targeted to top management [1/0]
Panel B: Weighted average premium OTM
in plans targeted to top management [1/0]
CEO ownership 2.090 2.088 2.069 2.068 2.300 2.305 2.318 2.322
(1.02) (1.01) (1.00) (0.99) (1.13) (1.12) (1.13) (1.12)
Non-state ownership control 1.311
n
1.312 1.319
n
1.319 1.478
n
1.474
n
1.472
n
1.468
n
(1.66) (1.63) (1.66) (1.63) (1.85) (1.82) (1.83) (1.80)
(0.88) (0.88) (0.89) (0.89) (0.78) (0.78) (0.77) (0.77)
State ownership [1/0] 0.265 0.265 0.265 0.265 0.369 0.369 0.370 0.369
(0.63) (0.63) (0.63) (0.63) (0.87) (0.87) (0.87) (0.87)
Firm size 0.082 0.082 0.086 0.086 0.074 0.074 0.071 0.071
(0.83) (0.82) (0.77) (0.76) (0.74) (0.73) (0.63) (0.63)
Tobins Q 0.079 0.080 0.078 0.078 0.085 0.084 0.087 0.086
(1.19) (1.06) (1.13) (1.00) (1.29) (1.14) (1.26) (1.10)
nn
1.418
n
1.406
nn
1.407
n
1.541
nn
1.530
nn
1.549
nn
1.539
nn
(2.17) (1.90) (2.12) (1.85) (2.32) (2.02) (2.29) (1.99)
(0.15) (0.15) (0.15) (0.15) (0.34) (0.33) (0.35) (0.34)
Cash ow-to-assets 3.494
n
3.495
n
3.460
n
3.460
n
3.925
nn
3.919
nn
3.955
nn
3.949
nn
(1.89) (1.87) (1.84) (1.82) (2.09) (2.07) (2.08) (2.05)
Prior stock return 0.216 0.216 0.224 0.224 0.075 0.073 0.068 0.067
(0.46) (0.46) (0.48) (0.47) (0.16) (0.16) (0.15) (0.14)
Prior stock return missing [1/0] 0.396 0.397 0.392 0.392 0.411 0.409 0.416 0.414
(0.47) (0.47) (0.46) (0.46) (0.48) (0.48) (0.48) (0.48)
nn
0.144
nn
0.143
nn
0.143
nn
0.145
nn
0.144
nn
0.146
nn
0.145
nn
(2.34) (2.02) (2.32) (2.00) (2.39) (2.04) (2.38) (2.02)
Firm focus [1/0] 0.126 0.128 0.058 0.057
(0.40) (0.41) (0.19) (0.18)
Mature industry [1/0] 0.127 0.128 0.056 0.055
(0.40) (0.40) (0.18) (0.17)
Growth industry [1/0] 0.124 0.127 0.071 0.068
(0.25) (0.25) (0.14) (0.13)
Total risk 1.422 1.423 1.138 1.132
(0.66) (0.66) (0.53) (0.52)
Systematic risk 2.012 2.011 0.632 0.646
(0.26) (0.26) (0.08) (0.08)
(0.36) (0.36) (0.38) (0.37)
(0.91) (0.91) (0.89) (0.89) (0.81) (0.80) (0.81) (0.81)
Dividendprotected plan [1/0] 0.027 0.027 0.025 0.025 0.021 0.021 0.023 0.023
(0.10) (0.10) (0.09) (0.09) (0.07) (0.07) (0.08) (0.08)
PV/indexed plan [1/0] 0.164 0.164 0.165 0.165 0.132 0.132 0.131 0.131
(0.34) (0.34) (0.34) (0.34) (0.27) (0.28) (0.27) (0.27)
Log likelihood 80.938 80.938 80.935 80.935 80.044 80.044 80.042 80.042
Pseudo R-squared 0.171 0.171 0.171 0.171 0.180 0.180 0.180 0.180
Dependent variable [at 1 in %] 0.518 0.518 0.518 0.518 0.525 0.525 0.525 0.525
support the results reported in Table 7. The previously
documented negative relation between the likelihood of
granting premium stock options and rm protability is
further strengthened by the results in Table 8, where all
coefcients of Cash ow-to-assets are negative and statis-
tically signicant in the reported specications. Finally,
in contrast to the results presented in Table 7, the results
in Table 8 suggest that a higher Capital-to-sales (lower
monitoring costs) reduces the likelihood of granting pre-
mium stock options to top management.
Finally, we test for a relation between the premium
and the vesting period. Bebchuk, Fried, and Walker (2002)
claim that in the US, at-the-money options are issued
irrespective of the vesting period. If stock prices are
expected to increase on average, then less managerial
effort would be required to make longer vesting-period
options end up in-the-money (such options would thus
constitute a royalty on the passage of time, and the
granting of such options would be an indication of
managerial power). Table 9, Panel A, reports descriptive
statistics for option programs with different numbers of
tranches. Tranches within a program are all granted at the
same date, but higher order tranches in our data
set always have a longer vesting period and often have a
different strike price (and usually have the same time-to-
maturity). As Panel A shows, option premiums are gen-
erally higher for tranches with longer vesting periods (the
differences are not signicant, however).
Higher option premiums for higher tranches might be
related not only to the increased vesting period, but also
to the higher time-to-maturity. Even if options with
longer maturities have higher exercise prices, they still
might be both more valuable at the grant date and
associated with a higher delta. That is, a higher strike
price does not guarantee that less value is distributed
through options with a longer maturity and vesting period.
We therefore construct a stronger test for the vesting-period
effect. Panel B of Table 9 reports results for pair-wise
comparisons of different tranches within the same option
programs. Only pairs with identical maturity but different
vesting periods (options from different tranches but for the
same rm and same program) are selected for comparison.
In xing maturity, we indirectly control for value because
given identical exercise prices, options with a shorter
vesting period should be at least as valuable as other options
(they can be exercised at the same time as others or earlier).
If exercise prices increase with the length of the vesting
period, then the results speak against the idea that options
with longer vesting periods would be more generous.
Panel B shows that all pair-wise average differences
are positive, that is, option premiums increase with
vesting periods even when maturity is kept constant.
When the sample size is larger (tranche 1 vs. tranche 2,
77 observations), the average difference in premiums is
signicant at the 1% level (a t-value of 2.48). The same
holds for the difference between the premiums for all
pairs with identical maturities (180 observations and a
t-value of 2.71). These results contradict the managerial
power hypothesis, and support the optimal contracting
view in the sense that when no institutional factors drive
granting at-the-money options, the shareholders will
grant more out-of-the-money options when their vesting
periods are further in the future to avoid rewarding
simply for the passing of time.
5.4. Additional tests related to comparability to
other markets
The results presented in this paper are based on data
for a small Nordic market (Finland). An important ques-
tion is whether these results also convey information
about the design of executive option programs in other
markets where additional tax and accounting rule-based
distortions do not apply.
When we compare Finland with the US and Continen-
tal Europe (see Section 4.1), we note that in terms of
ownership concentration, Finland lies in between the US
and Continental Europe but that there is much more
cross-sectional variation in ownership concentration in
Finland as compared with typical Continental European
countries.
To test for potential differences in the setting of the
exercise price that stem from different ownership struc-
tures, we split the sample into two approximately equal
parts (70 versus 71 observations) with regard to Non-state
ownership control. The average premium is larger in rms
with a non-state ownership share that is below average
(0.1155 vs. 0.0918), but the difference is insignicant.
Many authors argue that shareholder orientation in
Finland has increased over time.
21
Again, splitting the
sample into two parts based on the timing of the program
launch reveals that the premium has increased (i.e., the
latter options programs are more out-of-the-money,
though not signicantly so, with average premiums grow-
ing from 0.0699 to 0.1134) in the more recent subperiod
(19952001). Larger programs (i.e., programs with
potentially more powerful incentives) also have a higher
premium (0.1218 vs. 0.0851), but the difference is not
signicant. These results do not support the perception
that moving towards more shareholder orientation and
more powerful incentive compensation would lead to
decreased premiums and to programs with more at-the-
money options as in the US.
Benchmarking with respect to a peer group is found to
be widespread in recent empirical studies of compensa-
tion: see, e.g., Bizjak, Lemmon, and Naveen (2008) and
Faulkender and Yang (2010). The latter nd that rms
appear to select highly paid peers to justify their CEO
compensation. It is possible that executive option prac-
tices in Finland would also be inuenced by some impor-
tant peer group(s), especially because the number of
listed rms is small and a dominating advisory rm
(Alexander Corporate Finance) is involved in tailoring
21
See, e.g., Tainio and Lilja (2003) and Yl a-Anttila, Ali-Yrkk o, and
Nyberg (2004). There have also been several institutional changes during
our time period, supporting such a perception. These include the
increased foreign ownership (much of which is US-based) since 1993,
when restrictions for foreign ownership were abolished, new corporate
governance codes with features from the US or U.K., Finlands entry to
the European Union 1995, and many nancial market reforms increased
the competition and market orientation of Finnish rms.
most of the executive compensation programs. Different
peer groups might create signicant variation in option
premiums across industries, for example, or rms with a
foreign listing might follow different, more US-inuenced
standards. We test for the effect of peer groups both by
looking at the cross-sectional variation across industries
and by partitioning the sample based on foreign owner-
ship or international listing.
Although we do nd some variation across industries,
the average premium is always positive and is only below
Table 9
Analysis of option premiums vs. vesting periods.
Panel A reports (for each tranche) the average premium, standard deviation, time-to-maturity, and vesting period for options from programs with
different numbers of tranches. In Panel B, the analysis is restricted to different matched pairs (within a program) with identical time-to-maturity but
with different vesting periods, and potentially different premiums, since coming from different tranches. We report the statistics for the average
difference in premiums (higher tranche premium minus lower tranche premium) and vesting periods for such pairs. The t-test tests for whether such
average differences in premiums are different from zero. Signicant t-values at the 1% level (one-sided tests) are denoted in boldface. Final sample 141
observations (see Table 1).
Panel A: Number of tranches in program
1st Tranche 2nd Tranche 3rd Tranche 4th Tranche 5th Tranche
1 Average premium 0.0804
(N50) Stdev premium 0.1747
Average time-to-maturity 6.2003
Average vesting period 3.0923
2 Average premium 0.1181 0.1325
(N49) Stdev premium 0.1443 0.1414
Average time-to-maturity 5.9479 6.2542
Average vesting period 2.4310 4.0886
3 Average premium 0.1007 0.1346 0.1449
(N28) Stdev premium 0.2132 0.2325 0.2805
Average time-to-maturity 5.8062 6.0919 6.2866
Average vesting period 2.2297 3.3614 4.3534
4 Average premium 0.1796 0.1978 0.1980 0.2162
(N11) Stdev premium 0.1705 0.1971 0.2404 0.2721
Average time-to-maturity 6.1333 6.1333 6.4062 6.4062
Average vesting period 1.8371 2.7925 3.7479 4.7938
5 Average premium 0.0004 0.0004 0.0004 0.0004 0.0004
(N3) Stdev premium 0.0006 0.0006 0.0006 0.0006 0.0006
Average time-to-maturity 6.4950 6.4950 6.8283 6.8283 6.9945
Average vesting period 1.1306 1.7973 2.7991 3.4658 4.2995
All programs Average premium 0.1036 0.1367 0.1485 0.1699 0.0004
(N141) Stdev premium 0.1721 0.1797 0.2611 0.2557 0.0006
Average time-to-maturity 6.0354 6.1976 6.3566 6.4967 6.9945
Average vesting period 2.5516 3.6326 4.0838 4.5092 4.2995
Panel B: Different tranches (but same program and time-to-maturity) compared
1st Tranche 2nd Tranche 3rd Tranche All pairs
2nd Tranche Average difference in premium 0.0161
St. error of mean 0.0065
t-Test, paired sample 2.48
Average diff. in vesting period 1.3299
N 77
3rd Tranche Average difference in premium 0.0389 0.0222
St. error of mean 0.0276 0.0222
t-Test, paired sample 1.41 1.00
Average diff. in vesting period 1.9844 0.9152
N 34 32
4th Trance Average difference in premium 0.0269 0.0080 0.0143
St. error of mean 0.0473 0.0352 0.0091
t-Test, paired sample 0.57 0.23 1.57
Average diff. in vesting period 2.8658 1.9180 0.9646
N 11 12 14
All pairs Average difference in premium 0.0215
St. error of mean 0.0079
t-Test, paired sample 2.71
Average diff. in vesting period 1.4844
N 180
10% in two industries, IT & Telecom (an average premium
of 7%), and Consumer Staples (8.3%). The standard devia-
tions for the premium are wide in each industry, and the
differences in means are not signicantly different from
each other at the 5% level.
In Finland, foreign ownership varies substantially
across rms, from mainly foreign-owned and US-listed
rms such as Nokia to smaller, mainly domestically
controlled listed rms. When dividing the sample in two
based on foreign ownership, we nd that rms with
foreign ownership above the median have slightly smaller
premiums (0.0952 vs. 0.1120), and the programs of rms
with a foreign listing, mostly in the US (40 observations),
are even smaller (0.0754). However, these differences are
far from signicant and dually listed rms also have
programs that are, on average, 7.5% out-of-the-money.
6. Summary and conclusions
This paper examines the relations between rm char-
acteristics and the design of stock option plans for Finnish
rms. Most existing empirical studies concentrate on the
incentive effects provided by equity-based compensation
(see, e.g., Yermack, 1995). However, the variation in
contract design is rather limited for US rms, especially
in setting exercise prices, which is caused by tax and
accounting considerations (Murphy, 1999). In contrast,
Finnish rms are not subject to the tax- and accounting-
induced restrictions in the design of stock option plans.
This paper thus extends the literature in the following
manner. We examine determinants of the two main
design attributes of stock option plans: the scope of stock
option plans and the setting of exercise prices. In practice,
these two variables are shareholders main considerations
when evaluating and approving compensation proposals.
To our knowledge, this paper is the rst to provide a
rigorous analysis of the factors that determine the exer-
cise price of stock options. We also study the relation
between the out-of-the-moneyness of the granted options
and the length of their vesting period when options are
granted from several tranches.
We nd that the scope of the option plan is inversely
related to Tobins Q. Several prior studies employ Tobins
Q as a measure of rm performance (see, e.g., Morck,
Shleifer, and Vishny, 1988). Under that interpretation, this
result suggests that poorly performing rms grant larger
stock option plans. Furthermore, we nd that the scope of
stock option plans is decreasing in the capital-to-sales
ratio of the rm and in rm size and is increasing in rm
risk. These results are consistent with traditional princi-
pal-agent theory, which argues that greater monitoring
costs/difculty should be positively related to the amount of
equity-based compensation (Holmstrom, 1979; Demsetz
and Lehn, 1985; Milgrom and Roberts, 1992), and with the
idea that management productivity increases less than
proportionally with rm size (Baker and Hall, 2004).
We nd that the size of the stock option premium (its
out-of-the-moneyness) is very negatively related to the prior
stock return. This result is consistent with two alternative
interpretations, one from the optimal contracting literature,
and one fromthe managerial power literature. Under optimal
contracting, using high-water mark features in compensation
contracts (i.e., introducing out-of-the-money options) may be
optimal especially in high volatility industries. The alternative
view would be that managers have more negotiating power
regarding the design of compensation in rms with greater
prior stock price performance, which leads to more in-the-
money options. Our additional tests do not allow us to
strongly distinguish between these two alternatives. How-
ever, the rst view is somewhat supported by the result that
total risk is signicant and positively related to the option
premium.
In different specications for the option premium and
for the likelihood of launching premium options, we also
obtain support for other variables related to rm prot-
ability (such as cash ow-to-assets). These specications
indicate that less (more) protable rms are more likely
to grant premium (discount, i.e., in-the-money) options.
Because we also nd support for variables that proxy for
monitoring costs, our results concerning the exercise
price setting support both optimal contracting variables
but may also be interpreted as lending some support to
the managerial power hypothesis.
Finally, we report a signicant positive relation
between the option premium (out-of-the-moneyness)
and the length of the vesting period. In a pair-wise
comparison of different tranches from the same option
programs (with the same maturity, but different vesting
periods), we nd that the average option premium always
increases with the length of the vesting period. When
comparing all pairs, the difference between tranche
averages is signicant at the 1% level. Our results provide
clear support for an optimal contracting view in the sense
that when no institutional factors drive shareholders
toward granting at-the-money options, they will grant
more out-of-the-money options when the vesting periods
are further in the future to avoid reducing the managers
incentives over time.
Appendix A. Variable denitions for key variables
The table describes our key variables. Financial state-
ment data are from the ETLA database, rm ownership
data are from P orssiyhti ot books and the Finnish Central
Securities Depository (from 1993 onwards), nancial
market data are from DataStream and the databases at
the Hanken School of Economics, and ESOP data are from
Alexander Corporate Finance Oy. We have hand collected
press releases regarding the decisions of stock option
plans to enable the exact specication of, e.g., the stock
option premium at the date of grant. The time period is
19872001, and the sample covers 141 stock option
plans (Table A1).
Appendix B. Summary of expected signs for key
explanatory variables
The table summarizes the expected signs for our key
variables in the models for the scope of the stock option
plan as well as the exercise price. The signs are mainly
derived using predictions from either the optimal con-
tracting literature or the managerial power literature.
Table A1
Variable Description
Stock option overhang The number of shares exercisable to the sum of shares exercisable and the number of all already
outstanding shares at the grant date
Call option value of option plan to market
value of equity
For options without dividend protection, we use the Merton (1973) model, i.e., adjust for expected
dividends using the annual dividend yield; otherwise, we use the B&S (1973) model. The risk-free rate is
the three-month money market rate (HELIBOR, later EURIBOR) at grant date. The stock return volatility is
the annualized standard deviation of daily stock returns over 250 previous trading days (with a minimum
of 60 observations as inclusion criteria)
The total value of stock options in one tranche is their B&S value (with tranche-specic parameters)
multiplied by the number of shares obtainable upon exercise of all options in that tranche
The total value of stock options granted to the market value of equity is obtained by summing the total values of
all tranches in a stock option plan and dividing this value by the market value of rm equity at grant date
Adjusted call option value of option plan to
market value of equity
To adjust for the fact that managers may not be able to fully hedge their position in the market, we use
the method suggested by Meulbroek (2001) to value the options held by the fully undiversied manager.
The method is based on a Sharpe ratio approach where the manager is expected to require the same risk-
adjusted return for the stock as for the market, and the spot price is adjusted according to this. In
calculating the adjusted stock price, we have assumed a market return of 15%, risk-free rate of 5%, and
market volatility of 30%. The option values are then calculated as above using the adjusted value for the
current stock price
Stock option premium (tranche specic) Calculated as (XS)/S, where S is the current stock price and X is as follows:
(1) If the exact exercise price of stock options is specied at the grant date, then this value is used as X. This
is the case in most stock option plans in the study
(2) If the exercise price is dened as the average exercise price during some specied time period in the
future, then the stock option is assumed to be granted at-the-money (XS)
(3) If the exercise price of the option is dened as the average exercise price during some time period in the
future plus a xed or percentage premium, then the option is assumed to be granted with an exercise
price equal to S plus the given premium
(4) If the exercise price of the option cannot be determined with certainty at the date of grant, but a certain
minimum exercise price is specied, then the option is assumed to be granted with that minimum
exercise price. This is the case in a number of performance-vested/indexed stock options
First tranche premium [(XS)/S], where X is the exercise price in the rst tranche of the option plan, and S is the stock price grant
date
Weighted average premium (option plan
specic)
A weighted sum of the stock option premiums of all tranches in an option plan. As weights, the ratio of
the number of shares obtainable upon exercise of an individual stock option tranche, divided by the total
number of shares obtainable upon exercise of all tranches, is used
The target group of the stock option plans Stock option plans are dened as being targeted to top management if that group is the sole target of the
plan. If stock options are also targeted to non-executive employees, the stock option plan is dened as a
broad-based plan
CEO ownership The fraction of shares held by the CEO of the rm. P orssitieto only records the 20 largest shareholders of
the rm. This variable thus takes the value of zero if the CEO is not among this group of shareholders
Non-state ownership control The fraction of all shares held by the three largest non-state shareholders
Institutional ownership, State ownership Indicator variables taking the value of one if a nancial institution or the state is among the three largest
shareholders, respectively, and zero otherwise
Foreign ownership The fraction of shares held by foreign investors at the end of the accounting period
Firm size The logarithm of the book value of assets
Tobins Q The market value of equity and the book value of total debt, divided by the book value of assets. Market
values of equity (annual) are obtained from KOP P orssiyhti ot manuals and Kauppalehti databases
Investment-to-capital Gross investment in xed assets during the accounting period divided by xed assets (book value of gross
plant, property, and equipment)
Long-term debt-to-assets The book value of long-term debt divided by the book value of assets
Cash ow-to-assets The ratio of EBITDA to the book value of assets
Free cash ow-to-assets The ratio of EBITDA less gross investment and total dividends to the book value of assets
Prior stock return The six-month (125 trading days) logarithmic stock return preceding the end of the accounting period
during the year before the launch of stock option plans
Capital-to-sales The ratio of xed assets (book value of gross plant, property, and equipment) to sales
Wages per employee The ratio of total labor costs to the average number of employees
Firm focus A dummy that takes the value of one if at least 60% of the rms annual sales are generated from one
specic industry segment according to ETLAs classication, and zero otherwise (diversied industry)
Mature vs. Growth Firm focus is further decomposed into mature and growth industries. See Table 2, Panel C, for which
industries are classied as mature vs. growth industries
Dividend-protected plans In dividend-protected stock option plans, exercise prices are adjusted (reduced) on the ex-dividend date
for the amount of dividend payments per share
Total risk The variance of daily stock total returns during the rms accounting period, using a minimum of 60 daily
stock returns as inclusion criteria
Systematic risk Estimated by a year-to-year market model regression based on daily stock returns, and calculated as the
squared beta multiplied by the variance of daily market index returns. As the market index, we have used
the Helsinki Stock Exchange (HEX) Portfolio total return index from 1991 onwards (to control for the
Nokia effect, this index has a single rm weight cap of 10%), and prior to that, the WI index calculated at
the Hanken School of Economics (for its suitability, see, e.g., Knif, 1988)
Unsystematic risk Residual variance from a year-to-year market model regression based on daily stock returns
7 denotes a case where there are theoretical argu-
ments for both a positive and a negative effect, whereas
? denotes a case where, based on theoretical arguments,
an expectation for the sign is hard to formulate for some
control variable that we still want to include in the model.
Prior stock return is only included in model II (Table B1).
References
Aggarwal, R., Erel, I., Williamson, R., Stultz, R., 2009. Differences in
governance practices between US and foreign rms: measurement,
causes, and consequences. Review of Financial Studies 22, 31313169.
Baker, G.P., Hall, B.J., 2004. CEO incentives and rm size. Journal of Labor
Economics 22, 767798.
Banerjee, S., Gatchev, V.A., Noe, T.H., 2008. Doom or Gloom? CEO Stock
Options after ENRON. Unpublished Working Paper. Nayang Business
School, University of Central Florida, and Oxford University.
Barca, F., Becht, M. (Eds.), 2001. The Control of Corporate Europe. Oxford
University Press, New York.
Bebchuk, L.A., Fried, J.M., Walker, D.I., 2002. Managerial power and rent
extraction in the design of executive compensation. University of
Chicago Law Review 69, 751846.
Bhattacharya, P.S., Graham, M., 2009. Institutional ownership and rm
performance: evidence from Finland. Journal of Multinational Finan-
cial Management 19, 370394.
Bizjak, J.M., Lemmon, M.L., Naveen, L., 2008. Does the use of peer groups
contribute to higher pay and less efcient compensation? Journal of
Financial Economics 90, 152168.
Black, F., Scholes, M., 1973. The pricing of options and corporate
liabilities. Journal of Political Economy 31, 637654.
Choe, C., 2003. Leverage, volatility and executive stock options. Journal
of Corporate Finance 9, 591609.
Cuny, C.J., Martin, G.S., Puthenpurackal, J.J., 2009. Stock options and total
payout. Journal of Financial and Quantitative Analysis 44, 391410.
Dahiya, S., Yermack, D., 2008. You cant take it with you: Sunset
provisions for equity compensation when managers retire, resign,
or die. Journal of Corporate Finance 14, 499511.
Demsetz, H., Lehn, K., 1985. The structure of corporate ownership:
causes and consequences. Journal of Political Economy 93,
11551177.
Dittmann, I., Yu, K.-C., 2009. How Important are Risk-taking Incentives in
Executive Compensation? Unpublished Working Paper. Erasmus
University and Shanghai University of Finance and Economics.
Faulkender, M., Yang, J., 2010. Inside the black box: the role and
composition of compensation peer groups. Journal of Financial
Gaver, J., Gaver, K., 1993. Additional evidence on the association
between the investment opportunity set and corporate nancing,
dividend, and compensation policies. Journal of Accounting and
Goetzmann, W.N., Ingersoll Jr., J.E., Ross, S.A., 2003. High-water marks
and hedge fund management contracts. Journal of Finance 58,
16851717.
Grossman, S.J., Hart, O.D., 1983. An analysis of the principalagent
problem. Econometrica 51, 746.
Guay, W., 1999. The sensitivity of CEO wealth to equity risk: an analysis
of the magnitude and the determinants. Journal of Financial Economics
53, 4371.
Hall, B.J., Murphy, K.J., 2000. Optimal exercise prices for executive stock
options. The American Economic Review 90, 209214.
Hall, B.J., Murphy, K.J., 2002. Stock options for undiversied executives.
Journal of Accounting and Economics 33, 342.
Hansson, M., Liljeblom, E., L ound, A., Maury, B., Pasternack, D., Rosen-
berg, M., 2002. Kannustinj arjestelm at sek a niiden toimivuus suoma-
laisissa valtionyhti oiss a ja valtion osakkuusyhti oiss a. Kauppa-ja
teollisuusministeri on tutkimuksia ja raportteja 2.
Himmelberg, C., Hubbard, G., Palia, D., 1999. Understanding the deter-
minants of managerial ownership and the link between ownership
and performance. Journal of Financial Economics 53, 353384.
Holmstrom, B., 1979. Moral hazard and observability. Bell Journal of
Economics 10, 7491.
Holmstrom, B., 1992. Comments. In: Werin, L., Wijkander, H. (Eds.),
Contract Economics. Blackwell, Oxford, pp. 211214.
Ik aheimo, S., Kontu, H., Kostiander, L., Tainio, R., Uusitalo, A., 2007.
Ylimm an johdon palkitsemisj arjestelmien toimivuus valtionyhti oiss a
ja osakkuusyhti oiss a. Valtioneuvoston kanslian julkaisuja.
Jensen, M.C., 1986. Agency costs of free cash ow, corporate nance, and
takeovers. American Economic Review 76, 323329.
Jensen, M.C., 1993. The modern industrial revolution, exit, and the
failure of internal control systems. Journal of Finance 48, 831880.
Jensen, M.C., Meckling, W.H., 1976. Theory of the rm: managerial
behavior, agency costs and ownership structure. Journal of Financial
Table B1
Variable I. Model for the scope of the stock option plan II. Model for the stock option premium
Based on optimal
contracting
Based on
managerial power
Other Based on optimal
contracting
Based on
managerial power
Other
CEO ownership
Non-state ownership
control

Institutional ownership
State ownership
Firm size ?
Tobins Q 7
Investment-to-capital
Protability (cash ow-to-
assets)

Leverage (long-term debt-
to-assets)

Prior stock return
Capital-to-sales
Firm focus
Total risk 7 7
Systematic risk 7 7
Unsystematic risk 7 7
Prior plan in effect ?
Broad-based plan ?
Dividend-protected plan ?
PV/indexed plan ?
Jin, L., 2002. CEO compensation, diversication, and incentives. Journal
of Financial Economics 66, 2963.
John, K., John, T., 1993. Top-management compensation and capital
structure. Journal of Finance 48, 949974.
Klassen, K., Mawani, A., 2000. The impact of nancial and tax reporting
incentives on option grants to Canadian CEOs. Contemporary
Accounting Research 17, 227262.
Knif, J., 1988. Tests for Market Model Instability, An Empirical Compar-
ison of Tests Using Recursive Residuals. Research Reports 18.
Swedish School of Economics and Business Administration, Helsinki.
Korkeam aki, T., Liljeblom, E., Pastenack, D., 2010. Tax reform and payout
policy: do shareholder clienteles or payout policy adjust? Journal of
Corporate Finance 16, 572587.
Lambert, R.A., Larcker, D.F., Verrecchia, R.E., 1991. Portfolio considera-
tions in valuing executive compensation. Journal of Accounting
Research 29, 129149.
Lamont, O.A., Polk, C., 2001. The diversication discount: cash ows
versus returns. Journal of Finance 56, 16931721.
Makinen, M., 2007. CEO Compensation, Firm Size and Firm Performance:
Evidence from Finnish Panel Data. Discussion Paper 1084. The
Research Institute of the Finnish Economy.
Mehran, H., 1995. Executive compensation structure, ownership and
rm performance. Journal of Financial Economics 38, 163184.
Merton, R.C., 1973. Theory of rational option pricing. Bell Journal of
Economics and Management Science 4, 141183.
Meulbroek, L.K., 2001. The efciency of equity-linked compensation:
understanding the full cost of awarding executive stock options.
Financial Management 30, 544.
Milgrom, P., Roberts, J., 1992. Economics, Organization, and Manage-
ment. Prentice-Hall, Englewood Cliffs, New Jersey.
Morck, R., Shleifer, A., Vishny, R., 1988. Management ownership
and market valuation: an empirical analysis. Journal of Financial
Murphy, K., 1999. Executive compensation. In: Ashenfelter, O.C., Card, D.
(Eds.), Handbook of Labor Economics, vol. 3b. Elsevier Science,
North-Holland, pp. 24852563.
Palia, D., 2001. The endogeneity of managerial compensation in rm
valuation: a solution. Review of Financial Studies 14, 735764.
Palmon, O., Venezia, I., 2009. Stakeholders Welfare and Executive
Compensation under Managerial Overcondence. Working Paper.
Rutgers University.
Rosen, S., 1992. Contracts and the market for executives. In: Werin, L.,
Wijkander, H. (Eds.), Contract Economics. Blackwell, Oxford,
pp. 181211.
Rosenberg, M., 2003. Stock Option Compensation in Finland: An Analysis
of Economic Determinants, Contracting Frequency, and Design.
Unpublished Working Paper. Swedish School of Economics and
Business Administration.
Ross, S.A., 1973. The economic theory of agency: the principals problem.
American Economic Review 63, 134139.
Rosser, B.A., Canil, J.M., 2004. Executive Stock Options: Evidence that
Premium and Discount Awards do Matter. Unpublished Working
Paper. University of Adelaide.
Sauer, M., Sautner, Z., 2008. Stock Option Repricing in Europe. Unpub-
lished Working Paper. University of Amsterdam.
Schaefer, S., 1998. The dependence of pay-performance sensitivity on
the size of the rm. Review of Economics and Statistics 80,
436443.
Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance.
Journal of Finance 52, 737783.
Smith, C.W., Watts, R., 1992. The investment opportunity set and
corporate nancing, dividend, and compensation policies. Journal
of Financial Economics 32, 263292.
Tainio, R., Lilja, K., 2003. The Finnish business system in transition:
outcomes, actors and their inuence. In: Czarniawska, B., Sevon, G.
(Eds.), Northern Lights: Organisation Theory in Scandinavia. Copen-
hagen Business School Press, pp. 6987.
Yermack, D., 1995. Do corporations award CEO stock options effectively?
Journal of Financial Economics 39, 237269.
Yermack, D., 1997. Good timing: CEO stock option awards and company
news announcements. Journal of Finance 52, 449476.
Yl a-Anttila, P., Ali-Yrkk o, J., Nyberg, M., 2004. Foreign Ownership in
FinlandBoosting Firm Performance and Changing Corporate Gov-
ernance. The Research Institute of the Finnish Economy (ETLA)
Discussion Paper 904.

What Determines Stock Option Contract Designs

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

What Determines Stock Option Contract Designs

Uploaded by

Copyright:

Available Formats

What determines stock option contract design?

You might also like