260
BANKING STANDARDS
of loans and discounts, must be allowed for in forecasting the
changes in the equation, 6—4=2.
From Table 149 it is seen that, in general, the larger the
ratios of total deposits to earning assets, the smaller the ratios
of gross earnings. Accordingly, in series in which these ratios
exceed the average, the gross earnings ratio tends to be something
less than 6%. The total expense ratios and ratios of total de-
posits are indifferently correlated. Accordingly, ratios of total
expense may be more or less than 4%, the assumed average. It
follows, therefore, that the net earnings ratio may be less than
2%. This conclusion seems valid, because, as indicated in Table
153, ratios of deposits and ratios of net earnings are inversely or
negatively correlated. That is, high or increasing ratios of de-
posits tend to be accompanied by low or by decreasing ratios of
net earnings. If this is true, then it appears likely, since ratios
of gross earnings are positively correlated with net earnings and
ratios of total expense indifferently correlated with them, that
the effect of relatively large ratios of deposits in lowering net
earnings exceeds that of relatively large ratios of loans and dis-
counts in raising them. And yet this causal sequence does not
necessarily hold, inasmuch as, under actual conditions, variables
of a variety of sorts are operating simultaneously to determine net
earnings. The only thing which, from the analysis, seems to be
certain is that for the experience to which it relates, the types
of correlation to which attention is called obtain—all variables,
except those correlated, being ignored.
In a word, conditions associated with relatively high or increas-
ing net earnings in terms of earning assets are relatively high
or increasing ratios of loans and discounts to earning assets,
and vice versa; those accompanying relatively low ratios of net
earnings to earning assets are relatively high ratios of total
deposits to earning assets, and vice versa.
The manner in which such causal relations probably extend
through gross earnings and total expense, the immediate dif-
ferentials determining net earnings, has already been indicated.
Attenuated as is this chain of causal sequence, and unsatisfactory
as are the data—relating as they do to groups of banks—by
4Tn this connection, the proportion of demand deposits and of time deposits to
the total is of significance. See discussion in Chapter X.