SERIES CORRELATED WITH DEPOSITS 191
others which may be observed from the tables in this chapter,
but the following generalizations may be of some service in this
respect.
1. Yearly district ratios of gross earnings and of total expense
are generally low, relative to their respective district seven-year
levels, when district ratios of demand deposits to total deposits
are high with respect to their levels, and generally decrease from
year to year when demand deposits increase. Inverse relations
obtain when demand deposits are low and when they are decreas-
ing. The same respective relations obtain when demand deposits
are measured in terms of earning assets.
2. Yearly district ratios of net earnings are generally high,
relative to their respective seven-year district levels and to their
respective yearly country levels, when district ratios of demand
deposits are high.
3. Ratios of gross earnings and of total expense to earning
assets are lowest in districts in which ratios of demand deposits
to total deposits are high and ratios of loans and discounts to
earning assets are low. Ratios of net earnings to earning assets
are lowest in districts in which ratios of demand deposits to total
deposits and of loans and discounts to earning assets are low.
Conversely, ratios of gross earnings and of total expense are
highest in districts in which ratios of demand deposits are low
and loans and discounts are high, while ratios of net earnings are
highest in those districts having high ratios of both demand de-
posits and loans and discounts.
The relations between demand deposits as ratios of total
deposits and gross earnings, total expense, and net earnings,
respectively, expressed in terms of earning assets, are shown by
districts individually and combined in Charts 42, 43, and 44.
The bases upon which these charts are drawn and the manner
in which they should be interpreted are described in Chapter
VIII. The norms and trends of each of the series separately
have already been summarized. The charts are designed to
show the similarity and the percentage amounts of the deviations
of the yearly ratios from their respective district levels, and the
nature and percentage amounts of change from year to year
between the respective correlated series. They supplement the
tables presented above, inasmuch as they show that whatever