Using the Multivariate Data Analysis Techniques on the Insurance Market

by Dedu, Vasile; Armeanu, Daniel and Enciu, Adrian
Published in Romanian Journal of Economic Forecasting
, 2009, volume 12 issue 4, 170-179

 Requires a PDF viewer such as Xpdf or Adobe Acrobat Reader


In the present financial  theory, we confront with complex economic  phenomena and activities which cannot be studied or  analyzed profoundly because of the  plurality of existing variables, ratios and  information. The economic, financial and  social activity carried on under crisis or economic  growth conditions registered year by year  a development of the products and  instruments in use. The complexity of the  economic area may be simplified through techniques  of multi-dimensional analysis. Such a method is the analysis of the principal  components which allows the decreasing of  the initial causal space dimension generated  by the functional links which are  established among the initial explanatory variables.  The dimension of this space is determined  by the number of explanatory variables  identified as causes of the economic phenomenon and the higher their number,  the more difficult it is to analyze the  initial causal space because the information  volume, the complexity of calculations,  the risk not to identify the contribution of each  variable to the creation of the initial  causal space variability and the decrease in the  initial variables significance in case  they would be inter-correlated grow. The  simplification of the initial causal  space means the determination of a change which consists  in transition from a space with a large number of variables to another one of  fewer dimensions, equivalent but on the conditions of keeping maximum information  from the initial space and maximizing the variability of the new space (called  principal space). Variables from the  principal space represent the principal components, they  are un-correlated and the vectors which define them have a unitary length.

Keywords: original variables, covariance matrix, eigenvalue, eigenvector, principal components, total variance, generalized variance, factor matrix, factor loadings, factor scores, classification 
JEL Classification:
G15, G22, C44