Risultati dell

1
Risultati dellindagine sul Rischio Paese
condotta con le metodologie MCDM e SOM

• PhD Francesca Bernè, ing. Mattia Ciprian
• francescab_at_econ.units.it
• mciprian_at_units.it

2
Introduction - Generation Models - Country data -
Ratios

• An overview of country risk
• Modelling financial crisis the generation
  approach
• Financial crises an historical perspective
• Country risk assessment methodologies
• Country data and ratios

3
Introduction - Generation Models - Country data -
Ratios

• First-generation model (Krugman, 1979)
• Fixed exchange rate budget deficit monetary
  expansion drop in reserves
• financial crisis devaluation
• Second-generation crisis model (Obstfeld, 1985)
• Unsustainable fixed parity current account
  deficit
• Portofolio deficit capital outflows reserves
  exhaustion
• Exchange rate depreciation

4
Introduction - Generation Models - Country data -
Ratios

• Third-generation crisis model (Krugman, 1997
  Radelet and Sachs, 1998)
• Weak financial intermediation institutions bad
  governance
• Speculative short-term capital flows debt
  overhang and official reserve drop
• Financial panic and bank liquidations

5
Introduction Generation Models Country data -
Ratios

• Second-generation adjusted model of
  self-fulfilling crisis (Williamson, 2002)
• Macroeconomic fundamentals in intermediate
  situation (growth, inflation, current account,
  budget, debt)
• Multiple equilibria depending on
• regional contanimation speculative attacks
  bad equilibrium default
• robust adjustement credibility good
  equilibrium sustainable debt servicing
  capital market access

6
Introduction Generation Models - Country data -
Ratios

• Other models
• Maturity mismatch
• Currency mismatch

7
Introduction Generation Models - Country data -
Ratios

• Historical perspective
• Mexican crisis (1994)
• Russian crisis (1998)
• Asian crisis (1997-1998)
• Argentina crisis (2001-2002)

8
Introduction - Generation Models - Country data -
Ratios

• Country data
• Sources
• UNDP
• WORLD BANK
• COFACE
• OECD
• UNCTAD
• ICRG
• IMF
• CIA
• ISAE

9
Country data

• Countries (all the world)
• Europa and CSI, Americas, Asia and Oceania,
  Nord Africa and Middle East, Sub-saharian Africa

52 Countries, 22 ratios, year 2004, Argentina
2001-2002 Source CIA, COFACE

• Emerging Market Countries, Developing Countries,
  Least Development Countries

27 Countries, 18 ratios, period 1983 - 2000 (18
years) Source ISAE
Introduction Generation Models Country data -
Ratios
10
Ratios
birth rate, death rate, debt external, exports
(variation ), imports (variation), GDP
purchasing power parity, GDP no purchasing power
parity, inflation rate, public balance/GDP,
growth, net migration rate, investment, reserve
foreign exchange gold, infantility mortality
rate, life expectancy at birth, fertility rate,
labour force, internet users, industrial
production growth rate electricity consumption,
electricity production, oil consumption
Introduction Generation Models Country data -
Ratios
11
Ratios
default SPs, default state t-1 SPs, GDP
billion , GDP growth rate, rate of inflation,
exchange rate in purchasing power parity,
average interest rate, exports, foreign direct
investment, imports, total interest payment,
international reserve, total external debt,
short term external debt, interest on short term
external debt, total debt service, long term
debt service, current account balance
Introduction Generation Models Country data -
Ratios
12
SUMMARY

• Introduction
• Tools MCDM
• SOM
• Results
• Conclusions.

13
Introduction Tools (MCDM SOM) - Results
Conclusions

• INTRODUCTION
• The increasing complexity of financial problems
  over the past decades has driven analysts to
  develop and adopt more sophisticated quantitative
  analysis techniques furthermore, in the last
  years, is growing the opinion that the criterion
  to guide financial decisions has to be
  multidimensional.

14
Examples

• M. Doumpos, C. Zopounidis Assessing financial
  risks using a multicriteria sorting procedure
  the case of country risk assessment. OMEGA,
  vol.29, no. 1, 97-109, April 2000
• C. Zopounidis, M. Doumpos Multi-criteria
  decision aid in financial decision making
  methodologies and literature review. Journal Of
  Multi-Criteria Decision Analysis, 2002
• C. Zopounidis, M. Doumpos On the Use of a
  Multi-Criteria Hierarchical Discrimination
  Approach for Country Risk Assessment. Journal Of
  Multi-Criteria Decision Analysis, 2002
• G. Deboeck T. Kohonen Visual Explorations in
  Finance with self organizing maps,
  Springer-Verlag, 1998
• G. Deboeck Data mining with Self Organizing
  Maps a practical application, internet, 1998

Introduction Tools (MCDM SOM) - Results
Conclusions
15
Introduction Tools (MCDM SOM) - Results
Conclusions
MCDM

• Decision has inspired reflection of many thinkers
  since the ancient times. The great philosophers
  Aristotle, Plato, and Thomas Aquinas, to mention
  only a few names, discussed the capacity of
  humans to decide and in some manners claimed that
  this possibility is what distinguishes humans
  from animals.
• Classically, for example in economics, it is
  supposed that preference can be represented by a
  utility function assigning a numerical value to
  each action such that the more preferable an
  action, the larger its numerical value. Moreover,
  it is very often assumed that the comprehensive
  evaluation of an action can be seen as the sum of
  its numerical values for the considered criteria.
  Let us call this the classical model. It is very
  simple but not too realistic J. Figueira, S.
  Greco, M. Ehrgott Multiple Criteria Decision
  Analysis STATE OF THE ART SURVEYS Kluwer,
  2005.

16
Introduction Tools (MCDM SOM) - Results
Conclusions
MCDM

• Interactive methods
• Multi Attribute Utility Theory (MAUT)
• Outranking Methods.
• CODASID

17
Introduction Tools (MCDM SOM) - Results
Conclusions
CODASID

• This method attempts to generate a clear
  preference order for alternative designs.
• The basic concept is that the best action should
  have the shortest distance from an ideal design
  and the greatest from a negative ideal design.
• The inputs required by CODASID are
• a matrix containing the objects to be explored
  during the decision procedure
• a vector of weights expressing the relative
  importance of one attribute with respect to the
  others.

18
Introduction Tools (MCDM SOM) - Results
Conclusions
SOM

• The self-organizing map (SOM) is a new, effective
  software tool for the visualization of
  high-dimensional data. It implements an orderly
  mapping of a high-dimensional distribution onto a
  regular low-dimensional grid.
• Thereby it is able to convert complex, non-linear
  statistical relationships between
  high-dimensional data items into simple geometric
  relationships on a low-dimensional display.
• As it compresses information while preserving the
  most important topological and metric
  relationships of the primary data items on the
  display, it may also be thought to produce some
  kind of abstractions.
• The self-organizing map, T. Kohonen,
  Neurocomputing 21, 1998.

19
Self Organizing Maps
Introduction Tools (MCDM SOM) - Results
Conclusions
20
Results Evolution of developing countries

• We have obtained very encouraging results some
  of they are represented in figure.
• It is worth to note that every important fallout,
  historically occurred to a country, has been
  detected as a rank variation.

Introduction Tools (MCDM SOM) - Results (MCDM)
Conclusions
21
Results Evolution of developing countries
Introduction Tools (MCDM SOM) - Results (MCDM)
Conclusions
22
Results Ranking obtained from CODASID
Introduction Tools (MCDM SOM) - Results (MCDM)
Conclusions
23
Results SOM

• We have conducted two kinds of analysis on data
• factor study and individuation of local
  correlations in order to understand the logics
  joining the economics and social aspects of
  countries we trained a rectangular map with
  20x10 hexagonal nodes during 15000 cycles
• a clustering analysis based on credit risks, to
  produce a visual representation, and to verify
  the World Bank's classification according to the
  income level.

Introduction Tools (MCDM SOM) - Results (SOM)
Conclusions
24
Factor study

• We have obtained, as foreseen by us, some obvious
  relations as between Energy Consumption and
  Energy Production.
• The correlation is 0,963

Introduction Tools (MCDM SOM) - Results (SOM)
Conclusions
25
Factor study - Local correlation

• After all SOMs are very powerful in showing, in
  certain areas, the non-linear dependencies as in
  fig. The statistical correlation between the
  factors GDP Growth Rate and Industrial Production
  Growth Rate is very low 0,347.

Introduction Tools (MCDM SOM) - Results (SOM)
Conclusions
26
Cluster analysis
Introduction Tools (MCDM SOM) - Results (SOM)
Conclusions
27
Conclusions
As Meldrum said in a recent work a company needs
to examine the relationship between risk and its
businesses to make sure risk measures actually
help the company improve its business decisions
an additional risk is represented by country
risk. In our paper we show how SOM and MCDM could
be used as well tools to analyze this risk both
proved efficient in handling many data together
and producing a consistent classification of
countries according to their risk level. The
results have been compared with other recent
works.
Introduction Tools (MCDM SOM) - Results
Conclusions