Evaluation of Selected Methods of Classification for the Warsaw Stock Exchange

Article excerpt


The portfolio analysis allows a complex analysis of all the securities and it is connected with diversification of the portfolios risk. The problem that arises before the portfolios will be constructed and is connected with database of securities--what generally simplifies selection securities to portfolios. In a way of building database can be useful a taxonomic methods. The main aim of the paper is researching stability classifications for proposing methods and verification how different or similar they are. Such an approach in analysis of classification stability gives more information about researching companies and their financial or economic stability. It also gives information about the Polish capital market. In addition, such analyses are useful in making investment decisions, particularly in selection companies to portfolio. (JEL C10)


Securities analysis is an important element of investment. It is a complex and usually time-consuming process. Depending on the investment time horizon, the analysis may be carried out on the basis of appropriate technical or fundamental analysis tools. In long-term investments, it is the fundamental analysis methods and ratio analysis, in particular that are more useful. Moreover, more and more popular methods in securities analyses are taxonomic methods, which evaluate the quality of securities by means of synthetic analysis of economic and financial ratios. The combination of fundamental analysis methods with statistical, multidimensional methods can be described as follows:

1) Analysis of fundamental strength of companies, i.e., attractiveness of securities, with a synthetic list of economic and financial ratios as an element of securities analysis.

2) Application of economic and financial ratios as diagnostic variables in multidimensional analyses (e.g., to form databases, and possibility of their application to portfolio analyses).

3) Economic and financial ratio analysis, which allows evaluation of development of stock companies, and indirectly, the development of the capital market in Poland.

The methods used frequently to evaluate company fundamental strength are linear ranking methods, e.g., "taxonomic measure of attractiveness of investments" (TMAI), "generalized distance measure" (GDM), and "absolute development level ratio" (BZW). All these analysis elements focus on quantifiable economic and financial ratios. Generally, their levels directly affect the level of synthetic measures, and thus, the quality of classification. Moreover, a question arises, whether the selection of a classification method has a significant impact on the quality of classification and its stability in time. The issue is important, even due to the number of applicable methods. The paper suggests a need to discuss which of the analyzed methods is best in Polish conditions and how often the base of companies should be verified. These analyses concern, in particular, selected linear ranking methods, TMAI, GDM, BZW, calculated in two variants: with and without weights. Since the synthetic measures are based on certain economic and financial ratios as diagnostic variables, the quality of the original data is of great importance.

Research and Method Description

The paper analyzes the synthetic development measure TMAI, GDM, and BZW. To calculate the synthetic development measure the following formula was used [Tarczynski, 2002, pp. 101-11]:

TMA[I.sub.i] = 1 - [[d.sub.i]/[d.sub.0]], (i = 1, 2,...,n), (1)


TMA[I.sub.i] -- synthetic development measure for the i object,

[d.sub.i] -- distance between the i object and the pattern object defined with the formula:

[d.sub.i] = [square root of ([m.summation over (j=1)][w.sub.j] x ([z.sub.ij] - [z.sub.0j])[.sup.2])], (i = 1,2,...,n),

[d.sub.0] -- norm assuring that TMA[I.sub.i] reaches values ranging from 0 to 1:

[d. …