rr 
Hh 
11 
SERIES CORRELATED WITH NET EARNINGS 
THE net earnings of a bank, according to the accounting pro- 
cedure followed by the Federal Reserve Board in reporting the 
operating details of member banks, are the difference between 
the gross earnings and the total expense. That is, they are resi- 
duals, the amounts depending not upon the absolute size of the 
gross earnings and of the total expense but rather upon the dif- 
ference between them. For instance, a rate of 2%—relative to 
earning assets—may come about by virtue of a gross earning rate 
of 8% and a total expense rate of 6%, both expressed in terms 
of earning assets, as in the case of the banking experience of 
Dallas in 1922, or from a gross earning rate of 6% and a total 
expense rate of 4%, as, for instance, in the case of Philadelphia 
in 1925. Accordingly, if 8—~6=2, and 6—-4=—12, conversely, 
6+ 2=38, and 4 -}- 2=26. For a single bank, or for a group 
of banks treated as a unit, if gross earnings are large, other things 
being equal, net earnings are large, and if they are small, other 
things being equal, net earnings are small. Conversely, if total 
expenses are large, other things being equal, net earnings are 
small; if they are small, other things being equal, net earnings 
are large. 
But such simple relations do not obtain when averages, relating 
to any one or to all of the variables, are dealt with. This fol- 
lows because “the other things” which are supposed, as above, 
to be equal are unequal. For instance, large gross earnings are 
at times and at places associated with small, and at other times 
and places accompanied by large total expense, and vice versa. 
Conversely, large total expenses are at times and at places ac- 
companied by small and at other times and places by large gross 
earnings, and vice versa.! Accordingly, when data for individual 
banks, or for groups of banks taken as a unit, are combined with 
Em 
1 For a graphic summary of the relations between ratios of gross earnings and 
of total expense for member banks by districts, see Chart 33, page 130. 
52