The name is absent

TABLE 1. DISCRIMINANT FUNCTIONS WHICH
CLASSIFY FARM OPERATORS INTO
DEBT AND NO-DEBT GROUPS

Coefficients

Constant

Cost of equity (Kg)

Age of farm operator

Major source of income as a
percentage of total income

*.10 level of significance.

**.01 level of significance.

TABLE 2. VARIABLES USED IN DISCRIMINANT
ANALYSIS WITH MEANS AND BREAK-
EVEN VALUES

Variables

Mean Values

Debt No Debt Total

Group Group Sample

Break-even

Values

likelihood that a farmer would be classified as having
debt. Similarly, the negative coefficients for age and
the specialization variable indicates that increases in
these variables increase the likelihood that farmers
will have no debt.

The break-even values in Table 2 were calculated
with discriminant coefficients and sample means.
These break-even values indicate the specific magni-
tude of variables for which the discriminant function
is equated to zero if other variables are held at their
mean values. If a particular farmer had a mean value
for two variables and a value of the other variable
above this break-even value, he would be expected to
be in the group favored by the coefficient of the third
variable. For example, a farmer with 55.36 percent of
his income from one commodity source and an age of
55.81 years would have debt if his K_e were greater
than 4.05 percent.

Estimated break-even values for age and speciali-
zation are consistent with prior expectations. The
54.44 value for age is in the middle age range in
which farmers are generally considered to be in-
creasingly interested irr stability of income and
therefore lower financial leverage. Farmers with
income from one source being greater than 48.09
percent of total farm income would be specialists
more concerned with risk of financial leverage than
diversified farmers.

The break-even value for K_e of 4.05 percent is of
more concern. With interest rates in the range of
seven percent in 1972, a farmer would have to have a
marginal income tax rate of 45 percent for K_d to be
less than 4.05 percent. As this marginal tax rate was
for the taxable income bracket of $26,000-$32,000,
this estimate of the break-even value for K_e appears
to understate the minimum K_e for the average farmer
to introduce debt into his capital structure. For a
farmer to take out a new loan in 1972, his interest
charge might be seven to eight percent, so he would
have to have K_e almost that high unless his taxable
income were high. A possible explanation for this
result is that many of the sample farmers with debt in
their capital structure took out loans when interest
rates were much lower, say four to six percent. A K_d
based on a lower interest rate of the 1960s could be
lower than 4.05 percent for relatively low tax rates: a
marginal tax rate of 20 percent and an interest rate of
five percent would result in a K_d = 4 percent. Inter-
est rates increased rapidly in the early 1970s, but it
would take many years for a cross-sectional sample of
farmers to completely adjust their capital structures
to higher rates.

CONCLUSIONS AND IMPLICATIONS

This paper reports on research which adapted the
concept of weighted average cost of capital to
conceptualize the use of zero financial leverage
among many farmers. The concept was demonstrated
to be consistent with previous research on this issue.
An empirical model was developed which included
cost of equity, age and a specialization variable to test
applicability of the model for a sample of Georgia
farmers. A discriminant analysis of the sample cor-
rectly classified 96 of the 121 farms in the sample with
all coefficients having expected magnitudes and being
significant. Thus, the cost of capital Conceptprovidesa
useful theory to derive empirical models for analysis of
issues associated with capital structure of agriculture.

Several shortcomings of the analysis need to be
stressed. Since optimal capital structure is a long-run
concept, the particular empirical results are sensitive
to short-run phenomena. A cross-section sample with
several years of time series data would be more
appropriate than the one-year cross-sectional data
utilized in this study. Using only one year of data
would be appropriate during a period of stable
internal and external financial conditions for agricul-
ture. As such a situation is rare, perhaps the model
could be developed into an adaptive expectations
framework.

<<   <   2   3   4   5   6