Einladung zu zwei Seminarvortr=E4gen von
Prof. Adi Raveh
School of Business Administration and Statistics Department The Hebrew University of Jerusalem, Israel
THE GREEK BANKING SYSTEM: REANALYSIS OF PERFORMANCE
Abstract: Multiattribute Evaluation of Greek Banking Performance has been done recently by Zopounidis et.al. (1995). They used an ordinal regression analysis as well as an additive utility model to obtain final ranking of a representative sample of greek banks. In this paper we reanalyze their data by means of two multivariate analysis methods. One method is a nonmetric multidimensional saling type called Smalliest Space Analysis (SSA-I), which is designated to analyze variables (e.g., 7 attributes). The second method is called Co-plot. It is also a graphic display that is designated to analyze observations (e.g., 16 banks) and attributes simultaneously. Some further comments and findings are uncovered from our analysis.
key words: Banking Performance, Co-plot, Graphic Display, Smallest Space Analysis
Zeit: Montag, 6.5.1996, 13.00 Uhr Ort: Kleiner Sitzungssaal, =D6sterreichisches Institut f=FCr Wirtschaftsforschu= ng (WIFO) Arsenal, Objekt 20 1030 Wien
und
CO-PLOT DISPLAY TECHNIQUE AS AN AID TO THE PREDICTION OF CORPORATE BANKRUPTCY
Abstarct: CO-PLOT, a new graphic display method for multivariate data analysis, is introduced and used to aid the prediction of corporate bankruptcy. Co-plot maps the rows of a matrix in such a way that similar rows (observations) will be closely located on the map. The sample units are exhibited as n points and the variables as p arrows, relative to the same axis and origin. Co-plot enables the simultaneous study of observations and variables for a set of data, hence its name. This method is applied here to study similarity among companies, the correlation structure among financial ratios (variables), and the mutual relationships between the companies' observations and financial ratios. Co-plot yielded clearer findings for the data set analyzed previously by Altman for the same purpose of prediction.
key words: Dissimilarity, Maximum Correlation, Multidimensional Scaling (MDS).
Zeit: Dienstag, 7.5.1996, 16.00 Uhr Ort: HS 12, BWZ, Br=FCnner Stra=DFe 72, 1210 Wien
Peter Brandner Department of Economics email: peter.brandner@univie.ac.at University of Vienna voice: +43-1-40103-3371 Hohenstaufengasse 9 fax: +43-1-5321498 A-1010 Vienna, Austria ----------------------------------------------------------------- =========================================================================