|
||||
|
Table 2:
Variable definitions and theoretical justifications Click to view model
and click here to view part 2 and 3
of this table. Table 2, Part 1: The performance equation, EQ1 |
||||
|
Variable
name |
Definition,
theoretical justification and data source |
|||
|
Pi,t
Performance |
A
generic abbreviation for the financial performance measured at time t for
firm i. One of the following two specific measures is applied: a market-to-book
ratio and a standard measure of the return on total assets see below. |
|||
|
MTBi,t Market-to-book ratio |
MTB is
measured by: |
|||
|
RonAi,t
Return on assets |
For industrials it is calculated
as (Net Income before Preferred Dividends + ((Interest Expense on Debt - Interest
Capitalized) * (1-Tax Rate))) / Last Year's Total Assets * 100. For financials
it is different. For instance, for banks it is (Net Income before Preferred
Dividends + ((Interest Expense on Debt - Interest Capitalized) * (1-Tax
Rate))) / (Last Year's Total Assets - Last Year's Customer Liabilities on
Acceptances) * 100.I More details in Chapter 5, Section 3.1. Get
dissertation.
Data
source: Worldscope-Piranhaweb. |
|||
|
OWPi,t,d Piecewise managerial ownership |
Ownership
is measured either by total ownership by all officers and directors or by
insider ownership. The measurement details are described below in this table,
Part 2. To account for a possible
non-linear relation both of these ownership measures have been transformed
using a piecewise linear function with the ownership intervals: [0;0.5%],
[0.5;1%], [1;5%], [5;30%], and [30;100%]. More details about this
specification follow below this table. As an explanatory variable in the
performance equation it is intended to confirm one of the following two
hypotheses: The natural selection argument (Hypothesis 5) or the combined
incentive entrenchment argument by Morck et
al. (Hypothesis 3) as explained in Chapter 2, Section 2.3 and Section
2.5. Get
dissertation. |
|||
|
LnSALPi,t,c
Log of manager salary |
A generic
abbreviation for the natural logarithm of managerial compensation measured at
time t for firm i. One of the following two specific measures is applied:
annual salary by the sum of the two highest paid officers and annual salary
by the sum of the two highest paid directors (see below). In both cases the
measure includes the dollar value of fixed salary, bonuses, stock options and
other perquisites. |
|||
|
LnOFFSALPi,t,c
Sum of salary for the two highest paid officers million $ |
This is the
annual salary or remuneration of the company’s officers as reported in the
proxy statement at time t for firm i.IV Specifically, LnOFFSALP is
measured as the natural logarithm of the sum of the salary of the two highest
paid officers. Furthermore, to account for non-linearities this measure is
split by a piecewise linear function measuring the salary in the ranges from
0 to 1 million dollars, from 1 to 6 million dollars and from 6 million dollars
and above. More details below this table. It is hypothesized that financial
performance increases by increasing officer compensation for low levels of
compensation, but decreases for high levels. This scenario could be explained
by the argument that firms with higher officer compensation are more likely
to attract the most capable and dedicated people. However, at very high
levels of compensation the officer becomes too rich too quickly to care about
continuing to work hard to keep his/her job. Data source: SEC-Piranhaweb. |
|||
|
LnDIRSALPi,t,c
Sum of salary for the two highest paid directors million $ |
This is the
annual salary or remuneration of the company’s directors as reported in the
proxy statement at time t for firm i.IV Specifically, LnDirSALP is
measured as the natural logarithm of the sum of the salary of the two highest
paid officers. Also, this measure is split by a piecewise linear function in
the ranges from 0 to 1/2 million dollars, from 1/2 to 3 million dollars and
from 3 million dollars and above. More details below this table. The
associated hypothesis should be similar to those that apply for officer
salary, although the cut-point of the piecewise linear functions are expected
to be lower than for officer salary because directors typically are paid less
and work less than full time. Data source: SEC-Piranhaweb. |
|||
|
DSALi,t
Salary dummy |
This is a
dummy equal to one whenever the observation of managerial salary (officer or
director) is missing and zero otherwise. In regressions where this dummy is
used, the salary observations have been completed by adding the 25% quantile
salary whenever the salary observations are missing. The 25% value was
selected because it is believed to be a reasonably good approximation of the
true values of the missing observations. The coefficient to DSAL should be
insignificant to support the hypothesis that firms with substituted values
are no different in terms of performance than other firms. |
|||
|
Assetsi,t
Total assets |
Assets are total assets.
For industrials it is measured the standard way. For financials it is
measured differently. For instance, for banks it is the sum of cash, total
investments, premium balance receivables, investments in unconsolidated
subsidiaries, net property, plant and equipment and other assets.I
Data source: Worldscope-Piranhaweb. |
|||
|
LTDebti,t
Long-term debt |
Long-term
debt is used for the measurement of the firm’s leverage by dividing it by
total assets. Total assets are defined above and long-term debt represents
all interest bearing financial obligations, excluding amounts due within one
year.I Leverage is included in EQ1 because it is believed to be a
potentially important governance structure like ownership. Furthermore, the
squared leverage is included to identify possible optimal capital structures.
In particular, a significant negative coefficient of the leverage and a
significantly positive coefficient of the squared leverage could be reasoned
by an argument saying that firms with strong financial performance prefer not
to borrow and firms with high debt levels may perform better because it
provides high-powered incentives. More details follow below this table. Data
source: Worldscope-Piranhaweb. |
|||
|
MarkCapi,t Market capitalization |
MarkCap
is the year end market capitalization of all outstanding shares all types
included.I The size variable is used as a control variable in most
studies of performance and it enters EQ1 using a logarithmic transformation:
LN(MarkCapi,t). It is hypothesized to be positively correlated
with performance. When performance is measured by the market-to-book ratio
the positive correlation could be reasoned by a liquidity premium for large
firms. Independent of applied performance measure a positive correlation
could be reasoned by at least two other stories: 1) a survivor bias stating
that exceptionally well performing firms in general grow faster and therefore
are larger; 2) monopoly pricing stating that large firms in general are more
able to earn monopoly profits than small firms; and 3) earnings volatility
stating that small firms in general are less diversified than large firms and
therefore more often are turning out negative earnings. Data source:
Worldscope-Piranhaweb. |
|||
|
BETAi,t Stock market beta |
This is the standard measure of
market risk as described in Chapter 5, Section 4. Beta is based on between 23
and 35 consecutive month-end price percent changes relative to a local market
index.I Beta is included in EQ1 to control for differences in
financial performance that are the result of differences in beta. In
particular, high beta firms should deliver higher performance to compensate
investors for higher levels of non-diversifiable risk. More details below
this table. Data source: Worldscope-Piranhaweb. |
|||
|
PPECapi,t Total property plant and equipment net |
Total
property plant and equipment net represents total gross property, plant and equipment
less accumulated reserves for depreciation, depletion and amortization. Total
property, plant and equipment gross represents tangible assets with an
expected useful life of over one year, which are expected to be used to
produce goods for sale or for distribution of services.I Data source:
Worldscope-Piranhaweb. |
|||
|
CapExpi,t Capital expenditure |
This is the
capital expenditure from the cash flow statement. It represents the funds
used to acquire fixed assets other than those associated with acquisitions.I
Capital expenditure divided by total property, plant and equipment (see
above) is used in EQ1 to explain performance. One could argue that higher
capital expenditures to property, plant and equipment should be associated
with higher performance because such firms either are more willing to
undertake longer term investments and / or are more capable or positioned to
initiate profitable investment projects. Alternatively, if capital expenditures
are associated with depreciations that are higher than can be justified
economically, then the total asset values of firms with high capital
expenditures will be lower than they should be. This situation would create a
measurement bias for the applied market-to-book ratio (MTB) that would result
in a positive relation between MTB and CapExp to PPECap. However, this bias
is unclear with regard to return on assets because it affects both assets and
earnings downwards. Data source: Worldscope-Piranhaweb. |
|||
|
Salesi,t |
Net sales or revenues represent gross sales and other
operating revenue less discounts, returns and allowances.I Data
source: Worldscope-Piranhaweb. |
|||
|
OpeInci,t
Operating income |
This is operating income after
depreciation. It represents the difference between sales and total operating
expenses.I When divided by sales (see above) it measures the
profit or income per unit of sales. Ceteris
paribus higher profit rates should be associated by higher financial
performance. However, this is not the same as regressing financial
performance on financial performance (which would be absurd) because the
profit rate is only one of two basic components of financial performance that
also need to account for the capital used to create this profit. Data source:
Worldscope-Piranhaweb. |
|||
|
R&Di,t Research & development costs |
Research
and development expense represents all direct and indirect costs related to
the creation and development of new processes, techniques, applications and
products with commercial possibilities.I Although the primary
source was Worldscope, this item was collected from three databases in order
to boost the number of observations. More details below in Table 3, Part 2. Get dissertation. When
divided by total assets it measures the R&D intensity of the firm. If
performance is measured by q as defined above, a high R&D ratio should
also be associated by a high q because R&D intensive firms do not
capitalize R&D expenses on their balance sheet. However, it could also be
hypothesized that firms with higher R&D rates perform better because they
are more foresighted and have more potential for profitable inventions. Data
source: Worldscope/Extel/SEC-Piranhaweb. |
|||
|
DR&Di,t R&D dummy |
This
is a dummy equal to one whenever the observation of R&D is missing and
zero otherwise. In regressions where this dummy is used, the R&D
observations have been completed by entering zero whenever R&D is
missing. The coefficient to DR&D should be insignificant to support the
hypothesis that firms with substituted values are no different in terms of
performance than other firms. |
|||
|
Adveri,t Marketing costs |
The
marketing or advertising costs include expenses for postage, telephone, PR
and other similar marketing expenses as well as expenses in connection with
the publication of the annual and interim reports. This item was collected
from the Extel database.V When divided by total assets it measures
the advertising intensity of the firm. If performance is measured by q as
defined above, a high advertising ratio should also be associated by a high
q, because advertising intensive firms do not capitalize advertising expenses
on their balance sheet. However, perhaps it could also be hypothesized that
firms with higher marketing rates perform better because they are able to
sell more at higher prices than other firms. Data source: Extel-Piranhaweb. |
|||
|
DAdveri,t Marketing dummy |
This
is a dummy equal to one whenever the observation of Adver is missing and zero
otherwise. In regressions where this dummy is used, the Adver observations
have been completed by entering zero whenever Adver is missing. The
coefficient to DAdver should be insignificant to support the hypothesis that
firms with substituted values are no different in terms of performance than
other firms. |
|||
|
DIndust1i,t Three-digit industry dummies |
This
is three-digit industry dummies created from a four digit SIC code. The
classification contains (119-1) dummies. It is hypothesized that performance
is highly determined by the type of industries either because some industries
actually perform better at certain times but also because industry-specific
accounting practices that biases the measurement of performance. Data source:
SEC-Piranhaweb. |
|||
|
DExchangei,t Stock exchange dummy |
This
is a dummy for the stock exchange (or primary stock exchange) that the firm
belongs to. The classification contains (5-1) dummies. In particular, the New
York Stock Exchange, NASDAC National, NASDAC International, American Stock
Exchange or a category of other US stock exchanges.IV It is only
applicable for the full sample. Such dummies are hypothesized to be important
determinants of performance as a result of the different codes of conduct
that each exchange has implemented. Data source: SEC-Piranhaweb. |
|||
|
DIncorpi,t Dummy for country of incorporation |
A
new category of dummies fixing the country of incorporation. Seven dummies
(7-1) were created covering the following locations: |
|||
, where TotDebt
is total debt defined as all interest-bearing and capitalized-lease
obligations, or the sum of long- and short-term debt. MarkCap is market capitalization and Assets are total assets as defined below.I More
details in Chapter 5, Section 5.1.