386

BANKING STANDARDS

These detailed studies of the constant regression tendencies in
each group, as given by the constants K, and K,, all show the
predominating influence of gross earnings over total expense in
determining net earnings, and are consistent with the evidence
revealed in the constants E,” and E./. It seems worth while
briefly to generalize about these constant regression tendencies
by referring again to Table IV, and by noting the signs and
amounts of the constants K; and K, in the different groups.
In Groups 2 and 4, gross earnings ratios are below their 1924
average (see Table I). If they tend to increase the following
year—to regress toward their average of 6.0% —net earnings
must increase in banks in which total expense does not change.
The large positive values of K, for these groups represent the
increases in net earnings and furnish strong indication of regression
 of gross earnings to type. In Groups 1 and 3, gross earnings
are high in 1924. If they tend to decrease the following year,
net earnings must decrease in banks in which total expense did
not change. The negative values of K, for these groups represent
 the decreases in net earnings, and furnish evidence of regression
 in gross earnings.
In Groups 1 and 4, total expense ratios are above their 1924
average of 4.2%. If they tend to decrease the following year,
the result must be an increase in net earnings ratios for banks
in which gross earnings ratios did not change. The increase in
net earnings ratios is shown in the positive values of K, for these
groups. In Groups 2 and 3, total expense ratios are below their
1924 average. If they tend to increase the following year, the
result must be a decrease in net earnings ratios for banks in which
gross earnings ratios did not change. This decrease in net earnings
 ratios is shown in the negative values of K, for these groups.

APPARENT EXCEPTIONS AND CONTRADICTIONS IN TABULAR ANALY-SES
 EXPLAINABLE IN TERMS OF CONSTANT REGRESSION
TENDENCIES IN NET EARNINGS

In the light of the discussion of constant regression tendencies
in net earnings, it is possible to explain and interpret the meanings
of certain values obtained in the analyses in Chapter XVII, which
appeared to be exceptions or contradictions to the ruling tendencies
 there discussed.