- SelectedWorks

Transcription

- SelectedWorks
Wroclaw Univesity of Economics
From the SelectedWorks of Józef Z. Dziechciarz
2003
Applicability of the Discriminant Analysis
Framework for Polish Manufacturing Sector
Performance Assessment
Jozef Z. Dziechciarz, Wroclaw Univesity of Economics
Dorota Kwiatkowska Ciotucha, Wroclaw University of Economics
Urszula Załuska, Wroclaw University of Economics
Available at: http://works.bepress.com/jozefz-dziechciarz/74/
2
MER Revija za management
o
razvoj
MER [ournal for Management und En,u .
MER [ournal for Management and Dev.
ISSN
1408 - 9343
CDe
65·01
Letnik in śt. / Jahrgang und Nr. / Vol. and No.:
Programski svet / Herausgeberrat
5 (2003) 2
MER
Evrocenter, Gubno - Maribor,
November
/ Editorial Council;
Pro! dr. Janko Belak, Univerza v Mariboru in
~fER Evrocenter za management in razvoj,
Maribor in Gubno, Slovenija
Pro! dr. Vesna Brćic-Stipćević; Ekonomski
2003
Izdajatelji / Herausgeber / Publisher:
fakultet, Sveućiliśte u Zagrebu, Zagreb, Hrvatska
Abteilung fur Controlling und strategische
Pro! dr. Ićzef Dziechciarz,
Unternehmensfuhrung,
Institut fur
Wirtschaftswissenschaften, Pakultat fur
Wirtschaftswissenschaften und Informatik,
Universitat Klagenfurt, Klagenfurt, Osterreich
Akademia Ekonomiczna im. Oskara Langego we
Wrocławiu (Oskar Lange University of
Oconornics), Wrocłav, Polska
Ekonomski fakultet, Sveućiliśte u Zagrebu
Ekonomiczna im.Oskara Langego we
Wrocławiu, Wrocław, PoJska
(Faculty ofEconomics and Business, University
of Zagreb), Zagreb, Hrvatska
Ekonomski fakultet Rijeka, Sveućiliśte u Rijeci
(Faculty of Economics, University of Rijeka),
Rijeka, Hrvatska
Fakulta ekonomiky a manaźmentu
Economics and Management),
(Fakulty of
Slovenska
polnohospodarska univerzita v Nitre, Nitra,
Republika Slovakia
Institut fur Betriebswirtschaftslehre der Kleinund Mittelbetriebe an der Wirschaftsuniversitat
Wien, Wien, Osterreich
Institut fur Arbeitswissenschaft, RuhrUniversitat Bochum, Bochum, Deutschland
Instytut Inżynierii Zarzadzania (Institute of
Management Engineering), Wydział Informatyki
i Zarzadzania, Politechnika Poznańska, Poznań,
Akademia
Pro! Ing. Vladimir Gozora, PhD., Fakulta
ekonomiky a manaźrnentu, Slovenska
polnohospodarska univerzita v Nitre, Nitra,
Republika Slovakia
Pro! Ing. [aroslav Hornolka,
esC., Provoznć
ekonomicka fakulta, Ćeska zernedelska
univerzita v Praze, Praha, Ćeska Republika
Pro! Dr. Gyorgy Kadocsa, Szervezesi es Vezetćsi
Intezet, KeJeti Karoly Gazdasagi Kar, Budapesti
Musżaki Fóiskola, Budapest, Magyarorszag I
Hungary
Pro! Dr. Norbert Kailer, Institut fur
Arbeitswissenschaft,
Ruhr- Universitat Bochum,
Bochum, DeutschJand
Pro! dr. Stefan Kajzer, Univerza v Mariboru in
MER
Evrocenter, Maribor in Gubno, SJovenija
Pro! dr. Vinko Kandżija, Ekonomski fakultet
Rijeka, Sveućiliśte u Rijeci, Rijeka, Hrvatska
Pro! Dr. Dietrich Kropfberger, Abteilung fur
Controlling und strategische
Unternehmensfuhrung,
Institut fur
Wirtschaftswissenschaften,
Wirtschaftswissenschaften
Fakultat fur
und Inforrnatik,
Polska
Lehrstuhl Organisation und Personal, European
Business School Schlof Reichartshausen,
Pro! Dr. Josef Mugler, Institut fur
Oestrich -Winkel, Deutschland
MER Evrocenter za management
Mittelbetriebe an der Wirschaftsuniversitat
Universitat Klagenfurt, Klagenfurt, Osterreich
Betriebswirtschaftslehre
in razvoj
(MER
der Klein - und
Eurocentre for Management and Development),
Gubno, SJovenija
Provoznć ekonomicka fakulta (Faculty of
Economics and Management), Ćeska
zernedelska univerzita v Praz e, Praha, Ćeska
Wien, Wien, Osterreich
Republika
Szervezesi es Vezetesi Intćzet, KeJeti Karoly
Gazdasagi Kar, Budapesti Musżaki Fóiskola
Pro! dr. Zenon Wiśniewski, Wydział Nauk
(Institut of Organization and Management,
KeJeti Karoly Faculty of Econornics, Budapest
Polytechnic), Budapest, Magyarorszag I Hungary
Univerza v Mariboru (University of 1:aribor),
Maribor, Slovenija
Wydział Nauk Ekonomicznych
i Zarzadzania
(Faculty of Economic Sciences and
Management), Uniwersytet Mikołaja Kopernika
w Toruniu, Toruń, Polska
Pro! Dr. Jean-Paul Thornmen, Lehrstuhl
Organisation und Personal, Eurapean Business
School Schlof Reichartshausen,
WinkeJ, DeutschJand
Ekonomicznych
Oestrich-
i Zarzadzania, Uniwersytet
Mikolaja Kopernika w Toruniu, Toruń, Polska
Dr.-inź. Magdalena K. Wyrwicka, Instytut
Inżynierii Zarządzania, Wydział Informatyki i
Zarzadzania, Politechnika Poznańska, Poznań,
Polska
Recenzenti / Rezensenten / Reviewers:
Prof dr. Janko Belak, Univerza v Mariboru in
MER Evrocenter, Slovenija
Prof dr. Vesna Brćić-Stipćevu; Sveućiliśte u
Zagrebu, Hrvatska
Prof dr. hab. JózefDziechciarz, Akademia
Ekonomiczna im. Oskara Langego we
Wrocławiu, Polska
Doc. dr. Mojca Duh, Univerza v Mariboru,
Slovenija
Dr. hab. Marek Fertsch, Politechnika Poznanska,
Polska
Prof dr. Lovorka Galetić, Sveućiliśte u Zagrebu,
Hrvatska
Prof dr. [oźe Glogovśek, Nova Kreditna banka
Maribor, d. d. in Univerza v Mariboru, Slovenija
Dr. hab. Bohdan Godziszewski, Uniwersytet
Mikołaja Kopernika w Toruniu, Polska
Prof RNDr. [aroslav Havlićek, Csc., Ceska
zemćdelska univerzita v Praze, Cesl«i Republika
Prof Ing. [aroslav Homolka, CSc., Ceska
zernćdćlska univerzita v Praze, Ćeska Republika
Prof Ing. Jan Hron, DrSc., dr. h. C., Ceska
zernedelska univerzita v Praze, Ćeska Republika
Prof Dr. Gyiirgy Kadocsa,Budapesti Musżaki
.
Fóiskola, Budapest, Magyarorszag/ Hungary
Prof Dr. Norbert Kailer, Ruhr-Universitat
Bochum, Deutschland
Prof dr. Stefan Kajzer, Univerza v Mariboru in
MER Evrocenter, Slovenija
Prof dr. Vinko Kandżija, Sveućiliśte u Rijeci,
Hrvatska
Doc. dr. Iożica Knez-Riedl, Univerza v Mariboru,
Slovenija
Prof dr. Janko Kralj, Univerza v Mariboru,
Slovenija
Prof Dr. Dietrich Kropjberger, Universitat
Klagenfurt, Osterreich
.
Prof Dr. hab. Gabriela Łukasik. Akademia
Ekonomiczna im. Karola Adamieckiego w
Katowicach, Polska
Prof Dr. Thomas A. Martin, Fachhochschule
Kaiserslautern, University of Applied Sciences
FH-Campus Zweibrucken, Deutschland
Prof Dr. Iosef Mugler, Wirschaftsuniversitat
Wien, Osterreich
Prof. Dr. Christoph Midler, Universitat
Hohenheim, Deutschland
Prof dr. hab. Stanislaw Nowosielski, Akademia
Ekonomiczna im. Oskara Langego we
Wrocławiu, Polska
Prof Dr. Dr. h. c. f. Hanns Pichler,
Wirschaftsuniversitat Wien, Osterreich
Prof Dr. Hans Iobst Pleitner, Universitat
Sto Gallen, Schweiz
Prof dr. Danijel Pućko, Univerza v Ljubljani,
Slovenija
Prof Dr. Dietmar Ross], Wirtschaftsuniversitat
Wien, Osterreich
Prof Dr. Walter Ruda, Fachhochschule
Kaiserslautern, University of Applied Sciences
FH-Campus Zweibrucken, Deutschland
Prof dr. Marjan Senjur, Univerza v Ljubljani,
Slovenija
Doc. Ing. Miroslav Svatos, CSc., Ćeska
zernedelska univerzita v Praze, Ćeska Republika
Prof Ing. Karel Svoboda, CSc., Ceska zemedelska
univerzita v Praze, Ceska Republika
Prof Dr. [ean-Paul Thommen, European
Business School Schlofś Reichartshausen,
Oestrich -Winkel, Deutschland
Doc. Ing. Ivana Tichg, Ph. D., Ćeska zemedelska
univerzita v Praze, Ceska Republika
Dr. hab. Stefan Trzcielinski, Politechnika
Pomanska, Polska
Prof Ing. Jiii Tvrdon, CSc., Ćeska zemedćlska
univerzita v Praze, Ceska Republika
Prof dr. hab. Zenon Wiśniewski, Uniwersytet
Mikołaja Kopernika w Toruniu, Polska
Dr. inź. Magdalena K. Wyrwicka, Politechnika
Poznańska, Polska
Zaloźba / Verlag / Publishing house MER:
MER Evrocenter
Gubno 17, SI - 3261 Lesićno,
Slovenija / Slowenien / Slovenia
tel.: +386 (o)) 580 50 40
fax: +386 (o l3 580 50 41
e-mail: [email protected]
http.Y/www.mer-evrccenter.si/
Naslov urednlśtva / Redaktionsanschrift
Editorial and administrative
Mer Journal rur
Management und
Entwicklung erscheint
zweimal jiihrlich, im Mai
und November. Der Preis
pro Exemplar betragt (einschliefślich Steuer) fur
Abonnenten aus Slowenien
3.500 SIT und fur
Abonnenten aus dem
Ausland 23 €. Ihre
Abonnenten-Bestellung
nimmtdie
Marketingabteilung
unter
der Anschrift der Redaktion
entgegen. Bei diesen
BesteJlungen verringert sich
der Verkaufspreis um die
vereinbarte Rabattstufe.
/
office address:
MER Evrocenter
Zaloźba MER v Mariboru
Koroska 113 b, SI - 2000 Maribor
Slovenija / Slowenien / Slovenia
tel.: +386 (0)2 229 85 85
fax.: +386 (0)2 229 85 81
e-mail: publishingrś'mer-evrocenter.si
http.v/www.mer-evrocenter.si/
Odgovorni urednik / Chefredakteur
prof. dr. Janko Belak,
/ Editor-in-Chief:
SI - 2000
Maribor
Pomoćnlca odg. urednika / Stellv. Chefredakteurin
/
Assistant Editor-in-Chiefin:
doc. dr. Mojca Duh,
SI - 2351
Kamnica
Urednik- / Redakteur- / Editor-Consultant:
prof. dr. Stefan Kajzer,
Koordinatorica
/ Koordinatorin
SI - 2000
Maribor
/ Coordinator:
Marija Urśić
Lektoriranje
/ Lektoren / Readers:
Danijela Ravnjak (za SI)
Christiane Kordić (fur DE)
Barbara Żniderśić (for EN)
Oblikovanje / Gestaltung / Design:
Mat jaz Tornaźić
Tehnićna ureditev in realizacija / Technische
Technical realisation
Realisierung /
:
Bruno Żniderśić
Priprava za tisk / Druckvorbereitung
/ Preparation for print:
Repro studio OK, Maribor
Tisk / Druck / Printers:
Koda Press - Tiskarna Saje, Maribor
Copyright © MER Evrocenter, Gubno 2003
ISSN 1408-9343
MER Revija za management
in razvoj izhaja dvakrat
letno, maja in novembra.
Cena posamezne śtevilke
znaśa (vkljućno z davkom)
za naroćnike iz Slovenije
3.500 SIT in za naroćnike
izven Slovenije 23 €.
Abonentska naroćila
sprejema marketinśka sluźba
na naslovu uredniśtva, Pri
teh naroćilih bo prodajna
cena zniżana za
dogovorjeno stopnjo rabata.
Mer Journal for
Management and
Development is published
twice a year, in May and in
November. The price of
each magazine (ine. tax) is
3.500 SIT in Slovenia and 23
€ abroad, Applicalions for
annual subscriptions can be
sent to the Marketing
department at Editor's
address, Annual subscriptions will be given a red uetion.
Vsebina / Inhalt / Contents
Znanstveni
Poroćila
Banki
Wissenschaftliche
Beitrage
Berichte
Reports
Scientific Papers
SOCIAL
RESPONSIBILITY
5
OF A FAMILY
BUSINESS
DES KREDIT-
STEUERSYSTEMS
[ożica Knez-Riedl
2
BEDEUTUNG
FORDERUNG
FAMILIENUNTERNEHMEN
90
UND
FUR DIE ENTWICKLL~G
\"0_-
IN POLEK
~
Gabriela Eukasik, Teresa Famułska
VON KLEINEN
FAMILIENUNTERNEHMEN
GRUNDUNGS-,
6
IN DER
ENTWICKLUNGS-
UND
TAILOR-MADE
EDUCATION
FOR SMAli
ENTREPRENEURS
DBERGABEPHASE
[aroslav Havlićek, Jan Hron, Ivana Tidui __
100
Norbert Kailer
7
3
WISSENSMANAGEMENT
BEIM
GENERATIONENWECHSEL
IN KLEINEN
CROATIAN
FAMILY
Lovorka Galetić
82
DER
IM Kl.EIXEX
L>'-U
UNTER."r:IDiŁ-';
~
110
8
OF WORK
ASPEKTEN
Maria Dobisova, Hana Borsukowi
FAMILIENBETRIEBEN
ASPECTS
EINIGEN
MITTELSTANDISCHEN
Dietmar Rbssl
4
Zu
ARBEITSVERHALTNISSE
- FAMILY
CONFLICT
IN
ApPLICABILITY
ANALYSIS
ENTERPRISES
MANUFACTURI
120
OF THE DISCR~flXAXT
FRAMEWORK
FOR POUSH
G SECTOR
PERFO~TE
ASSESSMENT
Józef Dziechciarz,
Dorota Kwiatkowska-Ciotucha,
Urszula Załuska
150
Znanstveni ćlanki
Poroćda
Wissenschaftliche Beitriige
Berichte
Scientific Papers
Reports
DRUZBENA ODGOVORNOST DRUZINSKIH
5
PODIETII
Gesellschaftliche
Familienbetriebs
Verantwortung
des
Social Responsibilir of a Family Business
joźica Knez-Ried,
90
POSPESEVANJE USTANOVITVE, RAZVOJA IN
PREDAJE MALI H DRUZINSKIH PODJETIJ
Fordening von kleinen Familienuntemehmen
in der Grundungs-,
Entwicklungsund
Ubergabephase
Promotion ofSmall Family Enterprises in the
Start-up, Development and Tronsfer Phase
Norbert Kailer
3
~
128
Tailor-made Education for Smali Entrepreneurs
7
8
120
138
NEKATERI VIDIKI DELOVNIH RAZMERI) V
MALIH IN SREDN)E VELIKIH POD)ET)IH NA
SLOVASKEM
Zu einigen Aspekten der Arbeitsverhaltnisse
im
kleinen und mittelstandischen
Unternehmen
Some Aspects of Labour Re/ations in Smali and
Medium Sized Enterprises in Slovak Republic
Mdria Dobisovd, Hana Borsukovd
110
VIDIKI KONFLIKTA DELO - DRUZINA V
DRUZINSKIH POD)ET)IH NA HRV ASKEM
Ashkte der Familienarbeitskonflikte
in oatischen Familienbetrieben
Aspects ofWork - Family Conflict in Croatian
Family Ente;ferises
Lovorka Ga etić
Die hier veroffentlichten
Beitrage werden auf dem
Symposium MER 2003
prasenticrt.
MALIM PODJETNIKOM PRILAGOJENO
IZOBRAZEV AN)E
Spezifische Ausbildung der KMU-Untemehmer
Iaroslav Havlfcek, jan Hron, Ivana Tichd __
100
MANAGEMENT ZNAN)A PRl MEN)AVI
GE ERACI) V MALIH DRUZINSKIH POD)ETIIH
Wissensmanagement
beim
Generationenwechsel
in kleinen
Familienbetrieben
Knowledge Management at the Generation
Transfer in Smali Family Enterprises
Dietmar Ross!
The Importance of Credit and Tax System for the
Development of Family Businesses in Poland
Gabriela iukasik, Teresa Famulska
6
Ćlanki, ki so objavljeni v tej
MER Reviji,
so bili predstavljeni tudi na
simpoziju MER 2003.
POMEN KREDITNEGA IN DAVCNEGA SISTEMA
ZA RAZVOI DRUZINSKIH PODIETIJ NA
POLJSKEM
Bedeutung des Kredit- und Steuersystems
fur
die Entwicklung von Familienuntemehmen
in
Polen
144
UPORABNOST DISKRIMINANTNE ANALIZE
ZA PRESO)O USPESNOSTI PROIZVODNEGA
SEKTOR)A NA POL)SKEM
Anwendbarkeit
der Diskriminanzanalyse
fur die
Ermittlung des Erfolgs des Produktionssektors
in Polen
Applicability of the Discriminant Analysis
Framework for Polish Manufacturing Sector
Performance Assessment
f6zef Dziechciarz, Dorota KwiatkowskaCiotucha, Urszula Załuska
150
Articles, published in this
MER Iournal were
presented at the MER 2003.
MER
REVIjA
/ MER
5 (2003) 2:
JOURNAL
POROCILA
1tEPOaTS
/ BERJCHTE
Applicability of the Discriminant Analysis
Framework for Polish Manufacturing
Sector Performance Assessment
Józef Dziechciarz
Dorota Kwiatkowska-Ciotucha
Urszula Załuska
Introduction
1
ous advantages, including: robustness over time,
interpretability
and straightforwardness.
The
n recent years in Poland the demand for the
specific use of the discriminant function for the
information
manufacturing
I
deseribing
individual
situation
of the
industrie s or branches
branch is understood
(the
branches assessment may be for-
mulated as follow:
- Having the discriminant
as unit, group or class) is
function that was
rapidly growing. During these years - authors
estimated on the data from the given time
were working on the assessment of the Polish
period - the future situation of objects (e.g.
manufacturing
branches
result some papers
Compiegne
2001,
financial
indicators
manufacturing
performance.
GfKl Munich
branches
- Having the discriminant
(ASMDA
used
for
clustering
and
of
investors'
point
of view. The
objects
open
be
used
comparability?
analysis
to
The results
applied
assessment.
determined
branches
of discriminant
to the branches
situation
to lower
-
level (e.g.
down as the enterprise.
tion of the later task. This is due to severe problems with the data on low aggregation levels.
This task is becoming more and more important
was
in the context
solve
the
which requires enterprise rating. This may be
of the financial indicators
for
reasonable, comparable, stable methodology to
The
technique
by the
comparability
guarantee
belonging
(e.g. Groups)
The goal of the analysis here is the applica-
question remains - which financial indicators
may
classification
NACE
Classes) are classified. It may go as far
ordering in accordance with their attractiveness
from
function that was
estimated on the data from the higher level
Selected
2001).
were
enterprises) may be assessed.
As the
were presented
specific economic
attempt
choice
to
activity area. It has been
deterrnine
of Basel II recommendations
enterprise
rating
on measurable
assumed that if the use of the discriminant
qualitative and quantitative basis - common for
function leads to true classification of individual
the whole branch.
Prejem ćlanka /
Eingang des Beitrags /
Paper received:
objects (Branch) - it means that the financial
10·5·2003
construction
Recenzijska ocena /
Rezension /
Review evaluation:
raziskovalno poroćilo /
Forschungsbericht /
researcn report
may check in this manner,
indicators
used
for
discriminant
function
which indicators
have discriminative properties.
The elassie Fisher linear discriminant analysis technique was chosen because of its numer-
150
21
The Data
are comparable. Additionally, one
The Polish Central Statistical Office (cus)
collects the data describing the enterprise's eondition on the quarterly and yearly basis. In order
to assess branch condition
- the situation
of
JÓZEF
DZIECHCIARZ
DOROTA
individual firms belonging to this branch is to
KWIATKOWSKA-CIOTUCHA
URSZULA
debt ratio (long-term debt I equity), nominant
be examined. To cover all aspects of the enter-
(0,5);
prises activity the folIowing areas should be
4)
EFFICIENCY
ANALYSIS
-
cost ratio
X9
ineluded into analysis: sale's tendency; liquidity;
(prime cost of sales / sales), destimulant or des-
debt
timulant with threshold value
situation;
Corporate
efficiency
and profitability.
finance and managerial accounting
l; X10
fixed assets
utilization ratio (sales / fixed assets), stimulant;
productivity on one employee (sales / aver-
literature gives a variety of measures describing
Xll
the management quality (see [2,5]). The follew-
age number of employees), stimulant;
ing indicators have been chosen for the analysis
5)
PROFITABILITY
ANALYSIS
- X12
profit mar-
- the nature of the variable (stimulant, destimu-
gin on sales (net income / sales), stimulant or
lant, nominant)
stimulant with threshold value o;
and suggested nominal values
are stated:
l)
return on
X13
assets (net income / total assets), stimulant or
SALE'S
TENDENCY
ANALYSIS
xi the
-
stimulant with veto threshold value o; X14 return
dynamics of incomes from sale in fixed prices
on equity (net income / equity), stimulant or
from January
stimulant with veto threshold value o.
2002 -
chain base index - analog-
ous period previous year = 100%, stimulant;
2)
LIQUIDITY
ANALYSIS
-
xz current
The research was carried out on data gath-
ratio
ered according to the
For the evaluation
NACE.
(current assets / current liabilities), nominant
of the Polish manufacturing
sector performance
with recommended
the data for the branches
from the statistical
value range
[1,2 - 2,0]; X3
DR.
ZAŁUSKA
JÓZEF DZIECHCIARZ,
Professor, Head of the
Department of Eeonometries,
Wrodaw University of
Economics
ADDRESS:
Wrodaw University of
Eeonomies, ul. Komandorska
118/120,53-345 Wrodaw,
Po/and
Tel/fax: +48713680 359
E-mai/:
[email protected]/
MAIN
SCIENTIFIC-RESEARCH
FIELD:
econometru: modeling,
eeonometrie modeling with
variab/e and random
eoefficients, modeling on
qua/itative data and
mu/tivariate methods of data
analysis, small business
administration,
market
research, market data analysis,
macroeconomic data modeling,
macroeconomic data analysis,
business surveys,
manufaeturing branehes data
analysis, SMEs management,
quick ratio (current assets - inventories / cur-
reports col1ected by the Polish Central Statistical
SMEs situarion assessment,
rent liabilities), nominant
Office have been used. The yearly data from
infonnation needs
determination
value range [1,0 - 1,5];
tory utilization
X4
with recommended
finished goods inven-
ratio in days (finished goods
inventory / sales *
360),
destimulant;
X5
+ receivable turnover
D
for
92
groups and for
(Manufacturing)
181 elasses
ofSection
was used.
DR.
DOROTA
KWIATKOWSKA-
researcher and lecturer at the Department of
Forecasting and Economic
Analyses, Wrodaw University
of Eeonomies
CIOTUCHA,
cash
(financial means) cyele ratio in days (inventory
turnover
2000,
31
The analysis
- accountsIn accordance with the goal of the research
payable turnover), destimulant;
AODRESS:
3)
total
DEBT ANALYSIS
assets),
- X6
nominant
value range [0,57 - 0,67];
debt ratio (total debt I
with recommended
X7
debt-equity
(total debt / equity), nominant
mended value range
ratio
with recom-
[1,0 - 3,0]; X8
long-term
the discriminant
function
is constructed
in
order to predict the profit generation ability by
the enterprises. The discriminant
function eon-
structed is based on the data aggregated to the
level of Groups. The discrimination
criterion
Wrodaw University of
Eeonomies, Ul. Komandorska
118/120,53-345, Wrodaw,
Poland
Tel/fax: +48713680359
E-mail:
[email protected]
MER
MAIN
REVI)A
/ MER
SCIENTIFIC-RESEARCH
FIELD:
forecasting in enterprise, sale
forecasting, financial
forecasting, data analysis,
ciustering, nonhierarchical
5 (2003) 2:
JOURNAL
POROCJLA
dividing the whole Groups entity on two sub-
nating power are included. In the standard forward stepwise procedure - at the starting point
was the profit level. Depending
whether the profit was positive or negative - the
there is only one variable in the model. In the
Groups population
next step the next (one) variable is added. The
(92 in total) was divided
into two subpopulations.
manufacturing branches data
analysis, 5MEs management,
were:
DR.
URSZULA
choice criterion for the inclusion is the increase
of the Mahalanobis distance between subpopulations. In the standard backward stepwise pro-
of manufacturing
Groups with positive net profit in year 2000 -
cedure - at the starting point all variables are in
altogether 52 objects (Groups),
the model. In the next step one variable is omit-
nI
ZAEUSKA,
researcher and lecturer at the
Department of Forecasting and
Beonomic Analyses, Wrocław
University of Economics
Out of them there
no - subpopulation
SMEs situation-assessment,
determination
TS
populations
classification, macroeconomic
data analysis, business surveys,
information needs
RE."O
/ BERICHTE
-
subpopulation
of
manufacturing
ted. The choice criterion for the omitting is the
Groups with net loss in year 2000 - altogether 40
increase of the Mahalanobis
objects (Groups).
subpopulations.
For the discriminant
function construction
distance between
Here, both procedures lead to the same list
AODRESS:
Wrocław University of
Economics, Ul. Komandorska
118/120,53-345, Wrocław,
Poland
tel/fax: +48713680359
E-Mail:
[email protected]
MAIN
SCIENTIFIC-RESEARCH
the financial indices shown in section 2 were
of variables included in the discriminant
func-
used. Because of the adopted discrimination cri-
tion. These are: X9cost ratio, X3liquidity quick
terion (profit level) - the list of indicators was
ratio, X7 debt-equity
ratio, X5 cash (financial
modified in the way that the indicators describ-
means) cycle ratio. Variables are listed in accor-
ing profitability (i.e. variables X12,Xl3,Xl4) were
dance with their declining discriminative power.
eliminated.
Two versions of data were used for discrim-
FIELD:
forecasting in enterprise, sale
forecasting, financial
forecasting, business cyc/es
forecasting, data analysis,
clustering, nonhierarchical
dassification, macroeconomic
data analysis, business
tendency surveys,
manufacturing branches data
analysis, 5MEs management,
5MEs situation assessment,
information needs
determination
The variables that were used for the discriminant function construction
(i.e. variables xi -
XlI) the routine testing procedures
ducted. In particular
were eon-
the cross correlation
as
inant function estimation - raw and standardised. From practical point of view the discriminant function based on the raw data is more
convenient.
On the other hand the parameter
well as the mean equality and the normality of
estimates
the distribution
units. In contrast the discriminant function esti-
were examined. As the result
the observation occurred that only variable X6
mation
depend
on variables
measurement
based on standardised
observations
(debt ratio) is normally distributed. Researchers
show the weights each variable contributes
very frequently report, that most of the eco-
the discrimination
nomie variables are not normally distributed. In
tive impact comparison
spite of the fact that the variables are not nor-
criterion - high absolute value of the param eter
mally distributed
- all eleven variables were
to
criterion. It allows the relaon the discrimination
estimate indicates heavy impact of the variable
taken for further analysis. In case of variables X8
on the discriminative power of the discriminant
and XlI mean equality testing show the lack of
function.
significant
differences
between
means in two subpopulations
arithmetic
- this in tum leads
Equation (1) and (2) shows the discriminant
function estimation results for raw data (M(O))
to the conclusion that those two variables are
and standardised data (M(S)):
lacking of the discriminative features - they were
M(O) = -18.215x9 + 1.792X3-0.394
omitted in the analysis. For the reminder of the
x7 -0.006x5 +17.391
variables the cross correlation was examined. In
M(S) = -0.726x9 + 0.389x3 -0.260X7
most cases the correlation was low.
-0.216x5 +1.332
Using standard forward stepwise and back-
Standard
(2)
statistical evaluation
criminant
TICA package the discriminant
power assessment of the whole function as well
function
was
estimated. This results in specific, iterative vari-
function
requires
of the dis-
ward stepwise procedures offered by the STATIS-
discrimination
as the individual variables. For the discrimina-
able selection for the model. The task is that in
tive power assessment of the discriminant func-
the model only variables with strong discrimi-
tion the Wilks' lambda statistics l has been used.
152
Józef Dziechciarz,
ApPLICABILITY
OF THE
DISCU],{INANT
ANALYSIS
FRAMEWORK
FOR POLISH
Dorota Kwiatkowska-Ciotucha,
MANUFACTURING
SECTOR
Urszula Załuska:
PERFORMANCE
ASSESSMENT
150 - 154
I Disaiminant
TABLE l
function quality measures
Lambda Wilks'
F(1,87)
Significance level p
Tolerance value
X9
0.577
29.970
0.000
0.852
X3
0.459
6.176
0.015
0.768
X7
0.443
2.803
0.098
0.807
Xs
0.440
2.212
0.141
0.929
Source: Own computation.
Lambda statistics have values from the [o; l]
interval and its low value proves high discriminative power of the function. For the discriminative power assessment of the individual (k-th)
variable the Wilks' partial lambda statistics l.k
and tolerance
value have been used. Partial
lambda statistics measures the impact of the
individual variable on the composite value of
the statistics lambda l (how much will the value
of statistics l increase provided k-th variable is
driven out of the discriminant
function).
The
tolerance value was computed as (1-R2) ofthe kth variable with all other variabies included in
the model (the proportion
of variance that is
unique to the k-th variable).
For the obtained discriminant
function the
obtained Wilks' statistics lambda l
with corresponding
higher - out of 52 manufacturing Groups only
one was wrongly classified (i.e. 98.1% accuracy).
In the case of 40 manufacturing Groups with
negative financial result (net loss) the classification accuracy was lower i.e. 92.5%.
Fuli assessment of the discriminant function
quality requires that for validation a different
data set is used, different than analytical one i.e.
data set used for estimation of the function. As
the validation data set a population of 181 manufacturing Classes was used. It was known, to
which subpopulation
the individual object
(Class) was belonging. The subpopulation po consisted of 97 objects, and subpopulation pl
consisted of 84 Classes. Cross-validation on the
validation data set indicates additionally the
robustness and applicability of the discriminant
function. Classification matrix for validation
data set is shown in the TABLE3 •..
= 0.42880
statistics F (4; 87) = 28.973,
•• TABLE 3
P = 0.000). The additional model's characteristics are given in the TABLE2 ... They indicate
low cross correlation of financial indicators used
for the discriminant
function
construction
as
weli as the significance of used varia bles.
Classification matrix shown in the TABLE2
]to
]t1
Total
2
I Classification accutacy fur
analytical data set
Properly classified
objects (in %)
98.1
92.5
95.7
Source: Own computation.
3TP
• .
p=0.56 p=O..:!.L
51
3
3.
54
I 38
Total
accuracy for
validation data set
Properly classified
objects (in %)
86.6
83.3
85.1
]to
]tl
p=0.54 p=0.46
13
84
70
14
98
83
Source: Own computation .
•. illustrates the classification accuracy for the
analytical data (Group level, data used for function estimation). There were 95.7% properly
classified objects. For the objects with positive
financial result (net profit) this accuracy was
TABLE
]to
]tl
I Classification
I
I
Classification matrix shown in the Table 3
illustrates the classification accuracy for Class
level, data used for validation of the discriminant function. There were 85.1% properly classified objects. The number of false classified
objects was almost equal in both subpopulations. The value of discriminant function for
majority of false classified objects was close to
cut-off value (discriminant
function value
diriding objects into two subpopulations, in our
case cut-off value was equal=o.acy).
Although the discriminative power for the
.. . n data set was lower of some ten per-
153
8
MER
REVI)A
/ MER
5 (2003) 2:
JOURNAL
POROĆILA
centage points,
REPoltr.;
/ BERICHTE
the constructed
discńminant
function based on financial indicators proved its
applicability as a useful tool for prediction
of
profit generation ability. It may be used for each
aggregation level i.e. Group, Class or even enterprise.
there is improvement
potential additional exercise has been made. A
new discriminant
function has been construct-
ed. As the analytical data set the 181 objects
(Classes level) was used. The new discriminant
function consists of the same financial indicators as the previous one. Apart of the variables
X9, X3, X7, X5,
which were included in previous
function, there was only one additional indicator included
(X4
finished goods inventory uti-
ratio - with lowest discrimination
power). This new discriminant
function classi-
fies properly 864)10 objects. The improvement
in comparison with previous version of discriminant function is very small, only some 1.4%.
This means that there is no need to go to very
low aggregation
level while constructing
dis-
crimination function.
The further research requires analysis of the
discrimination
accuracy
for the individual
enterprises as well as the possibility of the profit
generation ability prediction.
154
Altman E. I. {I993]: Corporate financial distress and bankruptcy. A
complete guide to predicting and avoiding distress and profiting
from bankruptcy, Wiley, New York.
Brigham E. F., Gapenski LiC. {2000]: Financial Management, The
Dryden Press, Chicago.
Gnanadesikan R. {1997]: Methods for Statistical Data Analysis of
Multivariate Observations, znd ed., Wiley, New York.
Kwiatkowska-Ciotucha
D., Zal uska
Dziechciarz f. {2002] :
Manufacturing
Branches in Poland - A Classification Attempt. In:
Opitz O., Schweiger M. (eds.), Explanatory
Data Analysis in
Empirical Research, Springer, Berlin, PP.479-487.
Tyran M. R. {1992j: The Vest-Pocket Guide to Business Ratios,
u.,
To check whether
lization
Literature
Prentice- Hall, New York.
Izjava o poslanstvu in navodila avtorjem / Missionserklarung
Mission Statement and Guidelines for Authors
und Anweisungen fUr Autoren /
http://www.mer-evrocenter.si/
V MER Reviji objavljamo prispevke s
podroćja managementa in razvoja podjetij ter regij in drugih okolij, ki dosegajo
potrebno vsebinsko in jezikovno raven
obravnave in ki tudi formalno ustrezajo
zahtevam za uvrstitev med:
izvirne znanstvene ćlanke,
pregledne znanstvene ćlanke,
strokovne ćlanke,
raziskovalna poroćila, recenzije, kritike
in prikaze knjig.
Izpolnjevanje
formalnih
kriterijev za
objavo ugotavlja uredniśtvo, znanstveno
in strokovno
oceno prispevkov dajo
recenzenti, jezikovno oceno o prispevku
pa poleg recenzentov śe lektorji MER
Revije. Prispevki naj bodo napisani v
angleśkem jeziku ter opremljeni z naslovom, povzetkom in kljućnimi besedami v
angleśćini.
Avtor preda prispevek z vsemi prilogami
na raćunalniśki disketi ali CD-jU. Tekst
naj bo napisan v wordu za windows in
urejen v datotekah: tekst prispevka, povzetek in kljućne besede, vsebinski po udarki, podatki o avtorju oz. avtorjih. Tako
urejeno dokumentacijo
s priloźenirna
dvema izvodoma izpisov in portretno
sliko pośljete na naslov uredniśtva. Za
prispevke, posiane po elektronski pośti,
uredniśtvo ne zagotavlja zanesljivega prejema in zanje ne prevzema odgovornosti.
Reviji objavljeni prispevki niso
honorirani.
V MER
Natanćnejśa
navodila najdete
spletni strani
http./ /www.mer-evrocenter.si/
na nasi
Naslov uredniśtva /
Redaktionsanschrift /
Editorial office address:
MER Evrocenter
Zalożba MER v Mariboru
Koroska 113 b
SI - 2000 Maribor
Slovenija / Slowenien / Slovenia
Tel.: +386 (0)2 229 85 85
Fax: +386 (0)2 229 85 81 .
E-mail: [email protected]
http://www.mer-evrocenter.si/
Im MER Iournal werden jene Beitrage
veroffentlicht, die sich mit den Fragen des
Managements und der Entwicklung der
Unternehmen,
Regionen und anderen
Umgebungen
auseinandersetzen
und
dabei die notige inhaltliche als auch sprachliche Ebene erreichen, damit man sie
formai in eine der folgenden Kategorien
einreihen kann:
originelle wissenschaftliche Beitrage,
ubersichtliche wissenschaftliche
Beitrage,
Fachbeitragc,
Forschungsberichte, Rezensionen, kritische Auseinandersetzungen und
Buchbesprechungen.
Die Erfullung der formalen Kriterien fur
die Veroffentlichung der Beitrage stellt die
Redaktion fest, die wissenschaftliche und
fachliche Bewertung nehmen die Rezensen ten vor, sprachlich werden die Beitrage
- neben den Rezensenten - auch von den
Lektoren des MER Iournals bewertet. Die
Beitrage sollten in englischer Sprache verfasst
und
mit
dem
Titel,
der
Zusammenfassung
und den Schlusselwortem versehen sein.
In the MER [ournal we publish contributions from the field of management and
development of enterprises and regions
and other environments, which reach the
required eontent and language level and
formally fulfil editorial requirements and
which also formally suit the requirements
for placement among:
• original scientific papers,
• review scientific papers,
• professional papers,
• research reports, reviews, critics and
book presentations.
Implementation
of formal criteria for
publication of a contribution is carried out
by the editor, scientific and professional
evaluation of the contribution is done by
reviewers; language evaluation of the
contribution
is reviewed and checked.
Conttibutions
must be written in the
English language, furnished with abstract
and key words in the English language.
The author will submit the contribution
to the MER Evrocentre with all enclosures
on a disk or cd. The text should be written
in Word for Windows and saved on a disk
in the following files: Contribution,
Abstract and Key words, Content Points,
Data about the Author, Disk or cd should
be marked with the surname of the author
and title of the contribution and accompanied by rwo copies of the contribution
and all enclosures, as well as a portrai
JhOi:CXITa h, ~e editor =0;:
~:ee
~'e~;:o:.:o~
- ~~"-:..zcuxez:

Similar documents