212 M. Marozzi and L. Santamaria
do not usually have a strong background in statistics. From the practical point of
view, our method is more natural since it imitates what many analysts implicitly do in
practice by focusing on the most important aspects, discarding the remaining ones. It
is always readily comprehended, while principal components are often quite difficult
to be actually interpreted. From the theoretical point of view, a unique and very mild
assumption should be fulfilled for using our method: that financial ratios follow the
larger the better rule. We do not have to assume other hypotheses, that on the contrary
should be generally assumed by other dimension reduction methods such as principal
component (think for example about the hypothesis of linearity) or factor analysis.
Moreover, it is important to emphasise that, if one considers the first or second linear
transformation method, the composite indicator simplifying procedure may be applied
also to ordered categorical variables, or to mixed ones, partly quantitative and partly
ordered categorical, with the unique concern of how to score the ordered categories.
5 Conclusions
When a financial analyst rates a company, many financial ratios from its accounting
books are considered. By computing a composite indicator the analyst can analyse
different combinations of ratios together instead of sequentially considering each ratio
independently from the other ones. This is very important since ratios are generally
correlated. A quick and compact procedure for reducing the number of ratios at the
basis of a composite financial indicator has been proposed. A practical application to
the liquidity issue has been discussed. We ranked a set of listed companies by means
of composite indicators that considered the following liquidity ratios: the current
ratio, the quick ratio, the interest coverage ratio and the cash flow to interest expense
ratio. The results suggest that analysts should focus on the interest coverage ratio in
addressing the liquidityissue of the companies. By applying also principal component
analysis to the data at hand we showed that our dimension reduction method should be
preferred because it is always readily comprehended and much simpler. Moreover it
requires a unique and very mild assumption: that financial ratios follow the larger the
better rule. However, financial analysts should pay attention to the industry sector the
companies belong to. We suggest that financial analysts should group the companies
on the basis of the industry sector before applying our reduction procedure.
References
- Aiello, F., Attanasio, M.: How to transform a batch of simple indicators to make up a
unique one? In: Atti della XLII Riunione Scientifica della Societa Italiana di Statistica,`
Sessioni Plenarie e Specializzate, CLEUP, Padova, pp. 327–338 (2004) - Arboretti Giancristofaro, R., Marozzi, M.: A comparison of different methods for the
construction of composite indicators. In: Atti del IV Convegno Modelli Complessi e Metodi
Computazionali Intensivi per la Stima e la Previsione, CLEUP, Padova, pp. 109–114 (2005) - Fayers, P.M., Hand, D.J.: Casual variables, indicator variables and measurement scales:
an example from quality of life. J. R. Stat. Soc. A165, 233–261 (2002)