GROSS EARNINGS IN DISTRICT I 281

These tables show that (1) banks having high or low ratios
of gross earnings to earning assets have high or low ratios,
respectively, of total expense and of net earnings to earning
assets, and (2) the higher or lower the ratios of the first type,
the higher or lower those of the other types. To the first generalization
 there are no exceptions as the data are classified; to
the second, the two exceptions apply to the groups with few
frequencies.
But the detail in Table 159 and the averages in Table 160
refer to the experience of the entire number of member banks
in the two years 1924 and 1925. Space is not available in which
to present similar tables for all of the banks for the separate
years and for the various groups of banks for the individual and
combined years. Suffice it to say that the relations presented
above are repeated with substantial uniformity for each
yearly- and group-class. Indeed, the relations obtain irrespective
of the averages that are used as bases for computing the deviations,
 of the way in which the deviations are expressed—i.e., absolutely
 or relatively, and of the number of banks included—i.e.,
the entire membership or any part.

TABLE 160

CORRELATION OF DEVIATIONS OF RATIOS IN PAIRED SERIES—M EMBER
Banks, BosToN FEDERAL RESERVE DISTRICT, 1924 AND 1925
(Percentage Deviations from Respective Yearly Averages for the
Respective Volume-Groups. See page 270.)

Position

INDEPENDENT VaARIABLE—Ratioe of
(Giross Earnings to Earning As~~*~

Distance from Avera

Percent»
Cronne

“verage
-rcentage

Number |
of
3ank-Years


DEPENDENT VARIABLES—Net
Average Percentage

Tatal Expense to
Tarning
Aseetg

“Tet Earnings
o Earning
Assets

Total

Ahove

Ralow

10 and over
jo and under 40
10 and under 30
o and under 20
Inder 1a

Under 1
to and under 20
10 and under 30
jo and under 40
sn and aver

~f

id
-

SC
-*n Q

-— TX .0

*332 bark-years
t212 bank-veare