The selected ratios were, according to literature, the
measures of short-term solvency, long-term solvency,
profitability and efficiency of the firms. The empirically-
based classification was not fully equivalent to the a priori
classification (US firms). We found the following factors:
solvency, profitability efficiency and dynamic liquidity (Yli-
Olli and Virtanen 1985).
An interesting result was that we found a very similar factor
pattern when Finnish data were used. However, the factor
pattern was similar only when we used the first differences
of the value-weighted averages of the ratios. Certainly,
when the first differences of equal-weighted indices were
used, the classification pattern also appeared to be to some
extent acceptable as far as the Finnish data were concerned.
After that we analyzed, using transformation analysis, the
long-term stability and structural invariance of the factor
patterns obtained. The resulting factor pattern - based on
value weighted averages of the selected ratios (in the
difference form) - displayd very clear time series stability and
strong structural invariance between US and Finnish data.
On the other hand, the results gave evidence of considerable
instability and slight structural invariance when the
variables were equal-weighted indices. Equal-weighted
averages were especially sensitive both for outliers and for
the heterogeneity in the data. These results confirmed the
great importance of aggregation method in the financial
ratio analysis when we use aggregate data.
(The Finnish Journal of Business Economics, 2-
1986, 112-132)