Financial regulation: an attempt to regulate complexity

1
Financial regulation an attempt to regulate
complexity

• Imre Kondor
• Collegium Budapest and Department of Physics of
  Complex Systems, Eötvös University, Budapest
• International Workshop on Coping with Crises in
  Complex Socio-Economic Systems
• ETH Zurich (Switzerland), June 8-12, 2009

This work has been supported by the National
Office for Research and Technology under grant
No. KCKHA005
2
Contents

• Finance as a complex system
• The existing regulation
• Outlines of a future regulation

3
THE COMPLEXITY OF THE FINANCIAL SYSTEM
4

• The financial system consists of a large number
  of very heterogeneous agents, there is no
  symmetry, the mean-field model of the
  representative agent is completely wrong.
• It is not true that the players are all equally
  informed, equally rational, they all can process
  the available information in the same way, all
  are price takers.
• It is also not true that the system produces only
  small fluctuations about equilibrium, except for
  some significant exogeneous impacts.
• Instead, there are recurrent huge endogenous
  fluctuations. (E.g., Black Monday in 1987 was a
  35 sigma event!)
• Mainstream equilibrium theory provided the
  ideological basis for the neoliberal revolution
  that led to the deregulation of markets.

5

• There are complicated strong interactions between
  the components of the system (e.g. contractual,
  credit- or ownership relationships)
• Collective effects (bubbles, herding, exuberance,
  panic, etc.)
• No separation of scales, the distribution of
  wealth, or the market capitalization of firms is
  power-law-like
• The system is global, cannot naturally be cut
  into parts.
• It has a multiattractor structure with several
  metastable states (various financial, economic,
  or political regimes)
• It is sensitive to initial and boundary
  conditions, all kinds of control parameters

6

• The number of relevant variables is huge
• Financial time series are irreducible, cannot
  be reduced to a small number of variables, we
  cannot make an AI system learn them. They
  constitute a non-stationary, multivariate
  stochastic process that exists in a single copy
• Regression analyses in many (more and more),
  power law distributed explicatory variables (?) !
• Long (sometimes very short) range memory

7

• Mode slaving habits, traditions, laws,
  institutions
• Emergence all these coming about bubbles,
  crises and their consequences.
• Adaptation (e.g. to regulation), learning,
  self-organization.
• The system is conditioned by the information
  about it (the impact of the appearence of the
  Black-Scholes formula on liquidity), the system
  reflects upon itself
• Self-fulfilling prophecies (e.g. panics, Santa
  Claus rally, etc.)

8
SOME ORDERS OF MAGNITUDE
9

• Rank   Country   GDP (millions of USD)  
• World 60,689,812
•   European Union 18,394,115
• 1   United States 14,264,600
• 2   Japan 4,923,761
• China (PRC) 4,401,614
• Germany 3,667,513
• France 2,865,737
•   United Kingdom 2,674,085
•   Italy 2,313,893
•   Russia 1,676,586
•   Spain 1,611,767
•   Brazil 1,572,839
•   Canada 1,510,957
•   India 1,209,686
•   Mexico 1,088,128
•   Australia 1,010,699 Source IMF 2008

10
March2009
11
Controlling complex systems

• Is it possible?
• On which scale and which time horizon shall we
  optimize?
• Optimization is value driven

12
Some developments leading up to the present crisis

• Deregulation (after several decades of rigorous
  regulation)
• Explosive spread of derivatives (especially
  credit derivatives). Their total gross
  outstanding value reached 20 times the GWP! The
  players are circularly insuring each other, that
  is themselves, a complete nonsense.
• Among other things, capital adequacy regulation
  made the banks interested in securitisation.
• Congress, the FED, SEC and other regulators
  actively supported and encouraged the spread of
  OTC trading and off-the-balance-sheet items,
  thereby turning around a century old trend and
  making price discovery and reliable valuation of
  the new instruments impossible.
• This enabled Wall Street to fabricate the
  packages of opaque, uncontrolled and non-traded
  instruments on an assembly line.

13

• Because of the development of the originate to
  distribute business model in the mortgage
  lending market, banks have lost their incentive
  to investigate the credit-worthiness of
  borrowers.
• The originate to distribute model greatly
  contributed to the development of the system of
  perverted incentives (astronomical bonuses,
  compensation linked to short term performance).
• The frightening mass of colletarized debt
  obligations and structured investment vehicles
  created a totally opaque and infinitely complex
  shadow banking system.

14
Securitisation

• Huge quantities of claims put into tradeable, in
  principle homogeneous, risk-packages and sold to
  pension funds, foreign banks, China, etc.. The
  rationale is to distribute risk among the
  economic players, furthermore, to bring back the
  money tied down in property onto the market,
  thereby increasing liquidity.
• The rating of these packages is, however, almost
  impossible. (Gaussian copula the formula that
  brought down Wall Street)

15
Surge in Securitization
16

• By 2007 the total outstanding value of
  derivatives had reached the order of .
  (The net value is, of course, just a fraction of
  this the total Gross Product of the World is of
  the order of , but for the wide
  distribution of maturities clearing takes a long
  time.)
• Because of their complicated valuation,
  derivatives are typically off the balance sheet
  items.
• The overwhelming bulk of credit derivatives have
  been traded over the counter, the conditions
  stipulated are vastly different, impossible to
  compare (400 pages long contracts).
• Nobody knows how much of these toxic items can
  still be found in the portfolios of the various
  institutions, which is the main cause of
  suspicion and global lack of trust.

17
Is it possible to regulate such a system?
18
REGULATION
19

• Central bank, supervisors, etc., a very
  heterogeneous system
• It protects the interest of orphans and
  widows.
• Tries to ensure a level playing field for the
  players.
• Protects society as a whole from the collaps of
  the financial system, tries to mitigate systemic
  risk.
• Its main tool has been an obligatory minimum of
  own capital.

20
Why do banks need own capital?

• The fundamental difference between a bank and,
  say, a manufacturing company is the large
  leverage at the bank.
• Own capital is meant to provide a cushion against
  risk, against sustained loss.

21
Economic capital
Distribution of loss on e.g. credit portfolio
average (expectation value)
default probability (depends on target rating)
expected loss covered by interest rate spread
unexpected loss to be covered by own economic
capital
22
International agreements

• The first Basel Capital Accord (1988)
• Bank of International Settlements, Basel,
  www.bis.org
• Cook - Committee reviewed the practice of the
  40 best banks Cook-rate 8 - an arbitrary
  value.
• The Agreement was originally meant to bind the
  large banks of the G10 countries
  (US,UK,D,F,I,CH,NL,S,B,CN,J,L), by now it has
  essentially been accepted globally.
• The capital charge of 8 contributed to the
  depletion of the credit portfolios of the largest
  banks (securitisation/credit derivatives), the
  risk migrated into sectors unaccessible for
  regulation. An illustration of the law of
  unintended consequences.

23
The trading book regulation

• An amendment to Basel I (1996)
• Standard model
• Risk measures and the calculation of the capital
  charge prescribed by regulator
• Internal (own) model
• The bank may develop its own model for risk
  assessment, but must get it approved by the
  regulator
• The Value at Risk obtained from the internal
  model must be multiplied by at least a factor 3
  to calculate the capital charge

24

• Hungary adopted the corresponding EU Directive in
  2000. As it happened, I had to intimately
  familiarize myself with this piece of regulation,
  first at the stage of the preparation of the law,
  later in implementing it at a bank.
• With some colleagues we found that the regulation
  is full of inconsistencies concave risk
  measures, wildly fluctuating capital charges,
  etc.
• I. Kondor,A. Szepessy, and T. Ujvárosi Concave
  risk measures in international capital regulation
• in Risk Measures for the 21th Century, Ch. 4.,
  pp. 51-59, ed. G, Szego, John Wiley Sons (2004)
• A. Gábor and I. Kondor Portfolios with nonlinear
  constraints and spin glasses, Physica A274, 222
  (1999)
• I. Kondor Spin glasses in the trading book, Int.
  J. of Theor. and Appl. Finance, 3, 537 (2000)

25

• The Second Capital Accord
• The result of a long preparatory process
• Motivations
• Banks wanted a more flexible, risk-sensitive
  regulation.
• With the introduction of more differentiated
  capital charges, regulators wanted to encourage
  the spread of more advanced risk management
  practices
• Main goals
• to ensure stability of the individual
  institutions, as well as the whole system
• level playing field
• to maintain the average capitalization of the
  whole system at the level of 8 , but the banks
  applying advanced methods get a reduction (while
  the backward ones are penalized).

26
Structure of the new Accord
Three pillars
Market discipline
Minimum capital charge
Role of supervision

• Capital charge based on rating rating failed
  completely
• Supervisors got strongly enhanced discretionary
  power did not make use of it
• Market discipline (would have) meant increased
  transparency total darkness reigns

27

• Wide range of consultations, debates, impact
  studies, amendments since about 1998.
• The political arguments, lobbying, horse trading
  diverted attention from the essential questions.
• EU introduction in 2008-ban. (In the US in 2010)
• In the meantime huge, 3 to 5 years databases had
  to be created, the preparation lasted several
  years and cost fortunes.
• The regulation worked out to be complicated at
  the end.
• By the time it was introduced, the crisis had
  arrived.

28
There were dissenting voices right from the
beginning

• E.g. Jón Danielsson, Paul Embrechts, Charles
  Goodhart, Con Keating, Felix Muennich, Olivier
  Renault, and Hyun Song Shin An Academic Response
  to Basel II. (LSE Special paper, May 2001)
• They emphasized that financial risk is
  endogeneous (the metaphore of the Millennium
  Bridge), that VaR is not convex, it is procyclic,
  etc.

29
Systemic risk obtained very little attention

• Systemic risk is the risk of the collapse of the
  whole system of international financial mediation
• Such a collapse would be a disaster of historic
  proportion for civilization
• International regulation seeks to mitigate,
  indeed, to exclude this, but they believed to
  achieve this via ensuring the security of the
  individual shops.
• Neither regulators, nor the industry could
  imagine such an event could ever occur.
• The bail-out campaigns of 2008-09 aimed at
  preventing such a disaster.

30
DOES REGULATION STAND A CHANCE?
31

• In its present form the system is opaque,
  unknowlable and uncontrollable.
• The complexity of the system must be reduced,
  even at the price of reducing liquidity.

32
More concretely

• Trading must be driven back to the exchanges
  where prices can be discovered.
• Derivatives must be valued and should figure in
  the books and reports.
• The extent of securitisation must be limited,
  banks should be obliged to keep a substantial
  part of the credit packages.
• Leverage must be limited
• Regulation must become adaptive Transaction
  costs should grow with the trading volume in
  order to prevent bubbles from running away.

33
The report of the de Larosiere Committee

• Jacques de Larosiere high position French
  administrator
• 1978-87 director of IMF
• 1987-93 governor of the Banque de France
• 1993-98 president of EBRD
• The report analysing the crisis was submitted to
  the EU Commission in February 2009

34
Excerpts from their recommendations

• Recommendation 1 The Group sees the need for a
  fundamental review of the Basel 2 rules. The
  Basel Committee of Banking Supervisors should
  therefore be invited to urgently amend the rules
  with a view to
• - gradually increase minimum capital
  requirements
• - reduce pro-cyclicality, by e.g. encouraging
  dynamic provisioning or capital buffers
• - introduce stricter rules for off-balance sheet
  items
• - tighten norms on liquidity management and
• - strengthen the rules for banks internal
  control and risk management, notably by
• reinforcing the "fit and proper" criteria for
  management and board members.
• Furthermore, it is essential that rules are
  complemented by more reliance on judgement.

35

• Recommendation 3 Concerning the regulation of
  Credit Rating Agencies (CRAs), the Group
  recommends that
• - within the EU, a strengthened CESR (Committee
  of European Securities Regulators) should be in
  charge of registering and supervising CRAs
• - a fundamental review of CRAs' business model,
  its financing and of the scope for
• - separating rating and advisory activities
  should be undertaken
• - the use of ratings in financial regulations
  should be significantly reduced over time
• - the rating for structured products should be
  transformed by introducing distinct codes for
  such products.
• It is crucial that these regulatory changes are
  accompanied by increased due diligence and
  judgement by investors and improved supervision.

36

• Recommendation 4 With respect to accounting
  rules the Group considers that a wider reflection
  on the mark-to-market principle is needed and in
  particular recommends that
• - expeditious solutions should be found to the
  remaining accounting issues concerning complex
  products
• - accounting standards should not bias business
  models, promote pro-cyclical behaviour or
  discourage long-term investment
• - the IASB (International Accounting Standard
  Board) and other accounting standard setters
  should clarify and agree on a common, transparent
  methodology for the valuation of assets in
  illiquid markets where mark-to-market cannot be
  applied
• - the IASB further opens its standard-setting
  process to the regulatory, supervisory and
  business communities
• - the oversight and governance structure of the
  IASB be strengthened

37

• Recommendation 6 The Group considers that
• - Competent authorities in all Member States must
  have sufficient supervisory powers, including
  sanctions, to ensure the compliance of financial
  institutions with the applicable rules
• - Competent authorities should also be equipped
  with strong, equivalent and deterrent sanction
  regimes to counter all types of financial crime.
• Recommendation 7 Concerning the "parallel
  banking system" the Group recommends to
• - extend appropriate regulation, in a
  proportionate manner, to all firms or entities
• conducting financial activities of a potentially
  systemic nature, even if they have no
• direct dealings with the public at large
• - improve transparency in all financial markets -
  and notably for systemically important hedge
  funds - by imposing, in all EU Member States and
  internationally, registration and information
  requirements on hedge fund managers, concerning
  their strategies, methods and leverage, including
  their worldwide activities
• - introduce appropriate capital requirements on
  banks owning or operating a hedge fund or being
  otherwise engaged in significant proprietary
  trading and to closely monitor them.

38

• Recommendation 8 Concerning securitised products
  and derivatives markets, the Group
• recommends to
• - simplify and standardise over-the-counter
  derivatives
• - introduce and require the use of at least one
  well-capitalised central clearing house for
• credit default swaps in the EU
• - guarantee that issuers of securitised products
  retain on their books for the life of the
• instrument a meaningful amount of the underlying
  risk (non-hedged).
• Recommendation 10 In order to tackle the current
  absence of a truly harmonised set of
• core rules in the EU, the Group recommends that
• - Member States and the European Parliament
  should avoid in the future legislation that
• permits inconsistent transposition and
  application
• - the Commission and the level 3 Committees
  should identify those national exceptions,
• the removal of which would improve the
  functioning of the single financial market
• reduce distortions of competition and regulatory
  arbitrage or improve the efficiency of
• cross-border financial activity in the EU.
  Notwithstanding, a Member State should be
• able to adopt more stringent national regulatory
  measures considered to be domestically
• appropriate for safeguarding financial stability
  as long as the principles of the internal

39

• Recommendation 11 In view of the corporate
  governance failures revealed by the current
  financial crisis, the Group considers that
  compensation incentives must be better aligned
  with shareholder interests and long-term
  firm-wide profitability by basing the structure
  of financial sector compensation schemes on the
  following principles
• - the assessment of bonuses should be set in a
  multi-year framework, spreading bonus payments
  over the cycle
• - the same principles should apply to proprietary
  traders and asset managers
• - bonuses should reflect actual performance and
  not be guaranteed in advance.
• Supervisors should oversee the suitability of
  financial institutions' compensation policies,
  require changes where compensation policies
  encourage excessive risk-taking and, where
  necessary, impose additional capital requirements
  under pillar 2 of Basel 2 in case no adequate
  remedial action is being taken.

40

• Recommendation 12 With respect to internal risk
  management, the Group recommends that
• - the risk management function within financial
  institutions must be made independent and
  responsible for effective, independent stress
  testing
• - senior risk officers should hold a very high
  rank in the company hierarchy, and
• - internal risk assessment and proper due
  diligence must not be neglected by overreliance
  on external ratings. Supervisors are called upon
  to frequently inspect financial institutions'
  internal risk management systems.
• Recommendation 16 A new body called the European
  Systemic Risk Council (ESRC), to
• be chaired by the ECB President, should be set up
  under the auspices and with the logistical
  support of the ECB.
• - The ESRC should be composed of the members of
  the General Council of the ECB, the
• chairpersons of CEBS, CEIOPS and CESR as well as
  one representative of the European Commission.
  Whenever the subject discussed justifies the
  presence of insurance and securities supervisors,
  the Governor could choose to be represented by
  the Head of the appropriate national supervisory
  authority
• - The ESRC should pool and analyse all
  information, relevant for financial stability,
  pertaining to macro-economic conditions and to
  macro-prudential developments in all the
  financial sectors.
• - A proper flow of information between the ESRC
  and the micro-prudential supervisors
• must be ensured.