Impactul calitatii mediului nconjurator

1
Rating Culture Dimensions and Inside Rating
Piata Rosetti 4, Et. 4, Sector 2, Bucharest Tel
40-1-313.43.11, 313.43.12, Fax [phone]
2

• Our main purpose is to identify the determinants
  upon which a rating culture can be developed in
  an emergent economy.
• rating must become a science with an acknowledged
  economical standard, not an art surrounded by a
  veil of secrecy (1)
• its research objectives integrate a knowledge
  level compatible with the complexity and social
  utility of the underlying economic phenomena
• the scientific research base must be complemented
  with practical findings in order to accommodate
  the international processes oriented to the
  potential business / company estimation
• the human capital must be structured and
  supported in actual socio-economic frameworks in
  order to build a rating culture
• the local / global characteristic must
  accommodate specific environments in order to
  accelerate the main determinants of a rating
  culture

3
The rating process

• An abstract rating process exhibits the
  following pattern
• Credit events X Companies ? 0, 1 ? Rating
  classes, where
• the first map is a transition probability
  identified as PoD, the default probability map,
• the second map is a random variable identified as
  the rating allocation map.
• A Default Estimation associated with a credit
  decision call the rating process in order to
  obtain the expected loss predictions, i.e.
  mathematical expectations of dedicated stochastic
  variables. Here there are two complementary
  points of view
• the practical point of view, i.e. historical data
  projection of rating classes into PoD (2)
• the academic point of view, i.e. rational
  inference of rating classes based on PoD. (3)
• As an example, Moodys, SP, Fitch, a. o.
  represent rating classes by
• scoring systems in the so called default mode
  paradigm. KMV
• represent the rating classes for the valuation of
  an asset using time
• series only in the mark to model paradigm.

4
The rating process

• The operative representation of a rating process
  contains two categories of explanatory variables
  corresponding to (4)
• objective facts (firm size, financial data and
  analysis, a. o.)
• subjective facts (quality of management,
  organizational structure, a. o.)
• used two estimate the quantitative loss
  characteristics of the abstract rating process
• Probability of Default
• Loss in Event of Default
• Expected Loss
• Distribution of Loss Experience
• However, the main process is supported by a
  social mechanism in an economic environment,
  mediated by communication and social learning, by
  which the objective informational content is the
  argument base for subjective driven perceptions
  (5)
• Consequently, if someone intends to estimate the
  potential of a business, some sound prediction
  models in mathematical form must be built.

5
The kernel of rating research mathematical
modeling

• The classical financial ratio methods are
  backward looking, ignore non-linear effects and
  must be adjusted in a static manner.
  Consequently, in order to create a benchmark (for
  private firm models for example) with standard
  requirements
• conceptually sound with respect to dynamical
  features
• empirically validated on representative data
• actively used to manage risk by controlled
  forecasting
• comparable across institutions
• someone needs an unifying framework integrating
  different levels of complexity which the rating
  process exhibits.

6
The framework of mathematical modeling RSCC

• Random systems with complete connections (6)
  allow us to take into account
• irreversibility
• past dependence (hereditary and memory effects)
• stochastic and determinist dependence
• non-linearities in algorithmic form (recurrent
  and recursive)
• otherwise stated, this framework generalizes the
  markovian dependence in its main fractures.
  Nowadays, this modeling concept covers
• fractal and scaling framework
• chaos in dynamical systems
• iterated functions systems
• automata
• genetic algorithms
• ARMA - type models
• a. o.
• being an unified approach (7).

(1989) (1992) (1994) (1996) (2000) (2002)
7
An illuminating argument / example

• The RSCC in a generalized form was recently
  applied in mathematical finance to the
  forecasting problem of interest rates series (8).
  Using some sound concepts and results from this
  theory, but also a numerical method and three
  data sets,
• the performance of the forecasting procedure
• the performance of a classical ARMA procedure
• were compared. The new method is nearly
  uniformly better than the traditional ARMA
  technique.
• As a consequence, taking into account such a
  result and the public models from other subjects
  in mathematical finances (9) together with the
  coverage of this framework, some concrete
  subjects are already identified.

8
Rating systems for SMEs

• Defining features of SMEs, namely
• strongly regional and locally oriented economic
  activities and consumers
• heterogeneous structure of SMEs statistical
  population
• dedicated assessment tools to estimate future
  stability and performance
• specific vulnerability factors
• monitoring of market environment at local level
• impose a more complex rating criteria in order
  to build a rating culture. The main components
  upon which a genuine social learning mechanism
  must be based integrates forecasting techniques
  with psycho-sociological perceptions influencing
  credibility and trust. Besides an entire class of
  subjectively oriented models (10), the human
  factor problem asks for a solution.

9
The Head Building Policy of the Rating Culture

• The main idea consists in a structured
  infrastructure of experts in which the
  improvement of knowledge necessary to control the
  future complexity works.
• From our point of view, the following
  qualifications are necessary
• Rating modeling qualification (not reduced to the
  purely mathematical aspects)
• Rating implementer
• Rating analyst
• Rating account
• in order to develop on solid ground the desired
  rating culture. The curricula for a clear
  description of these directions is in progress
  (11).

10
References and Comments

• (1) Banks that can calculate the default
  probability of their loans and the losses they
  would suffer in the event of a default would be
  allowed to use their internal systems to set
  regulatory capital, 1
• 1 W. McDonough (Chairman of the Basel
  Committee) American Banker 9(14) 2000
• (2) As an illustrative example of practical
  approach to this problem, cf 2
• 2 E. Falkenstein, A. Boral, L.V. Carty
  Risk Calc for private companies Moodys
  Default Model, Rating Methodology, May 2000
• (3) The following references belong to the
  succession series of this subject or a
  combination of both viewpoints
• 3 R.C. Merton Theory of Rational Option
  Pricing, Bell Journal of Economic and
  Management Science 4, 1973, pp 141-83
• 4 J.W. Wilcox A Simple Theory of
  Financial Ratios as Predictors of Failure,
  Journal of Accounting Research 9(2), 1971, pp
  389-395
• 5 E.I. Altman Financial ratios,
  discriminant analysis, and the predictions of
  corporate bankruptcy, Journal Finance 23,
  1968, pp 589-609
• 6 E.I. Altman et.al. ZETA Analysis A New
  Model Identity Bankruptcy Risk of
  Corporations, Journal of Banking and Finance 10,
  1977, pp 29-54

11
References and Comments

• (4) The only known public references exposing the
  structure of the methodology and operative
  representation of internal rating are
• 7 W.F. Treacy, M.S. Carey Credit Risk
  Rating at Large U.S. Banks, Federal Reserve
  Bulletin 11, 1998, pp 897-921
• 8 M.S. Carey Credit Risk in Private Debt
  Portfolios, The Journal of Finance 53(4),
  1998, pp 1363-87
• 9 J.P. Krahnen, M. Weber Generally
  Accepted Rating Principles A Primer Journal
  of Banking and Finance 24, 2000 pp
• (5) The social mechanism idea goes back to many
  contributors belonging to classical sociological
  schools. In Romania Harets 1910 contribution and
  its development in Gustis sociological school or
  Xenopols causality theory are the classical
  references. The American Sociology comes into
  this subject with Mertons essay On sociological
  Theories of the Middle Range (1967). The
  importance of this fact consists in a definite
  framework for the rating process which can not be
  reduced to an economical viewpoint some
  contemporary considerations are exposed in
• 10 P. Hedström, R. Swedberg (eds) Social
  mechanisms an analytical approach to social
  theory, Cambridge University, Press, 1998

12
References and Comments

• (6) The first RSCC ever studied is about what is
  nowadays called an OM chain, after two Romanian
  mathematicians Octav Onicescu and Gherghe Mihoc,
  who defined this concept in the middle 1930s and
  called it a chain with complete connections. The
  modern theory of RCC is part of M. Iosifescus
  mathematical work see
• 11 M. Iosifescu, S. Grigorescu Dependence
  with Complete Connections and its
  Applications, Cambridge University Press,
  Cambridge, 1990
• A valuable econometric approach using RSCC from
  am international viewpoint is due to
• 12 O. Onicescu, M.C. Botez Incertitude
  and economical modeling. Informational
  econometry, Ed. Stiintifica si Enciclopedica,
  Bucuresti, 1985 (in roumanian)
• (7) This property of a mathematical modeling
  framework seem to be desirable even for the
  practitioners cf 2 or other valuable public
  papers of SP, L.P. Morgan, Fitch, a.o.
• (8) Using the RSCC framework, this example is
  exposed in
• U. Herkenrath, A. Rundolph Forecasting of
  interest rate series, May 2002 (working
  paper)

13
References and comments

• (9) As an example Jarrow-Turnbul 13 model based
  on a markonian approach corresponding to the
  classical Elfvings 1937 imbedding problem (cf
  14 p. 246) can be generalized.
• 13 R.A. Jarrow, D. Lando, S.M. Turnbull A
  Markov Model for the Term Structure of Credit
  Risk Spreads, The Review of
• Financial Studies , 10(2), 1997, pp 481-523
• 14 M. Iosifescu Finite Markov Processes
  and their Applications, John Wiley Sons,
  NY, 1980
• Also, using some ideas from 12 concerning
  sound generalisations of Shannons entropy due to
  Guiasu 15 or Zapan 16 someone readily extends
  the framework of validity exposed in 17, 18,
  a.o.
• 15 S. Guiasu Mathematical Structure of
  finite random cybernetic systems, Udine,
  Italy, 1971
• 16 Gh. Zapan Mathematical modeling of
  human activities and evolutive process with
  some applications in psychology and pedagogy.
  Revista de pedagogie, 15(5), 1956 pp 21-39
  (in roumanian)
• 17 S.C. Keenan, J. R. Sobehart Performance
  Measures for Credit Risk Models, Moodys Risk
  Management Services, Research Report
  10-10-99 (1999)

14
References and Comments

• 18 J.R. Soberhart, S.C. Keenan, R. Stein
  Benchmarking quantitative default risk
  models A validation Methodology, Moodys
  Risk Management Services, rating
  Methodology, March 2000
• Other research directives will be exposed
  elsewhere.
• (10)  In fact it is necessary to integrate the
  QoL framework is an collective setting our
  generic idea is to extend the quality of life
  principles from sociological paterns to
  socio-economical organized groups. Otherwise
  stated the health of SMEs is deeply related with
  the well-being and welfare of individuals. In
  such a vision some
• classical models concerning
  socio-economical systems add value by a
  methodology complemented beyond the purely
  economical dimensions. Our inspiration points are
  
•  19 R.H.G. Field, D. Tjosvold, W.C. Wedley
  Constructive controversy, the Vrom-
  Yetton model, and managerial decision
  making, Journal of
  Occupational Behavior 7 (2),
  1986 pp 125-138
• 20 C. Zamfir
  Incertitude. A psycho-sociological
  perspective Ed.
  Stiintifica, Bucuresti, 1990
  (in roumanian)
•  

15
References and Comments

• 21 Ph. Yetton, M. Levy, Ph. Pow SMEs
  aligning IS and the strategic antext
  Journal of Information Technology 16, 2001, pp
  133-144
• (11)  Our program integrating the dimensions
  exposed at least, will be mature next year.