2FR04-InscoeFDIC Basel II

							        The Evolution to Basel II
       XBRL and the Basel II Capital Accord


Donald Inscoe
Deputy Director
Division of Insurance and Research
U.S. Federal Deposit Insurance Corporation




                XBRL International, Tokyo, Japan November 8, 2005
First Basel Accord

   The first Basel Accord (Basel I) was
    completed in 1988
    –   Set minimum capital standards for banks
    –   Standards focused on credit risk, the main risk
        incurred by banks
    –   Became effective end-year 1992
Reason for the Accord

   To create a level playing field for
    internationally active banks
    –   Banks from different countries competing for the
        same loans would have to set aside roughly the
        same amount of capital on the loans
1988 Accord Capital Requirements

   Capital was set at 8% and was adjusted by a
    loan’s credit risk weight
   Credit risk was divided into 5 categories:
    0%, 10%, 20%, 50%, and 100%
    –   Commercial loans, for example, were assigned to
        the 100% risk weight category
Risk-Based Capital

   The Accord was hailed for incorporating risk
    into the calculation of capital requirements
Capital Calculation

   To calculate required capital, a bank would
    multiply the assets in each risk category by
    the category’s risk weight and then multiply
    the result by 8%
    –   Thus a $100 commercial loan would be multiplied
        by 100% and then by 8%, resulting in a capital
        requirement of $8
Criticisms of the Accord

   The Accord, however, was criticized for
    taking too simplistic an approach to setting
    credit risk weights and for ignoring other
    types of risk
Risk Weights

   Risk weights were based on what the parties
    to the Accord negotiated rather than on the
    actual risk of each asset
    –   Risk weights did not flow from any particular
        insolvency probability standard, and were for the
        most part, arbitrary
Operational and Other Risks

   The requirements did not explicitly account
    for operating and other forms of risk that may
    also be important
    –   Except for trading account activities, the capital
        standards did not account for hedging,
        diversification, and differences in risk
        management techniques
Banks Develop Own “Capital
Allocation” Models

   Advances in technology and finance allowed
    banks to develop their own capital allocation
    (internal) models in the 1990s
   This resulted in more accurate calculations of
    bank capital than possible under Basel I
   These models allowed banks to align the
    amount of risk they undertook on a loan with
    the overall goals of the bank
Internal Models and Basel I

   Internal models allow banks to more finely
    differentiate risks of individual loans than is
    possible under Basel I
    –   Risk can be differentiated within loan categories
        and between loan categories
    –   Allows the application of a ―capital charge‖ to
        each loan, rather than each category of loan
Variation in Credit Quality

   Banks discovered a wide variation in credit
    quality within risk-weight categories
    –   Basel I lumps all commercial loans into the 8%
        capital category
    –   Internal models calculations can lead to capital
        allocations on commercial loans that vary from
        1% to 30%, depending on the loan’s estimated
        risk
Capital Arbitrage

   If a loan is calculated to have an internal
    capital charge that is low compared to the
    8% standard, the bank has a strong incentive
    to undertake regulatory capital arbitrage
   Securitization is the main means used by
    U.S. banks to engage in regulatory capital
    arbitrage
Example of Capital Arbitrage
   Assume a bank has a portfolio of commercial loans with the following
    ratings and internally generated capital requirements
     –   AA-A: 3%-4% capital needed
     –   B+-B: 8% capital needed
     –   B- and below: 12%-16% capital needed
   Under Basel I, the bank has to hold 8% risk-based capital against all of
    these loans
   To ensure the profitability of the better quality loans, the bank engages
    in capital arbitrage--it securitizes the loans so that they are reclassified
    into a lower regulatory risk category with a lower capital charge
   Lower quality loans with higher internal capital charges are kept on the
    bank’s books because they require less risk-based capital than the
    bank’s internal model indicates
New Approach to Risk-Based Capital

   By the late 1990s, growth in the use of
    regulatory capital arbitrage led the Basel
    Committee to begin work on a new capital
    regime (Basel II)
   Effort focused on using banks’ internal rating
    models and internal risk models
   June 1999: Committee issued a proposal for
    a new capital adequacy framework to replace
    the 1998 Accord
Basel II

   Basel II consists of three pillars:
    –   Minimum capital requirements for credit risk,
        market risk and operational risk—expanding the
        1988 Accord (Pillar I)
    –   Supervisory review of an institution’s capital
        adequacy and internal assessment process (Pillar
        II)
    –   Effective use of market discipline as a lever to
        strengthen disclosure and encourage safe and
        sound banking practices (Pillar III)
Pillar I

   In the United States, all banks that will be
    required to conform to the new capital
    standard will use the Advanced Internal
    Ratings Based approach (AIRB)
AIRB Approach Requirements

   Collect sufficient data on loans to develop a
    method for rating loans within various
    portfolios
   Develop a Probability of Default (PD) for
    each rated loan
   Develop a Loss Given Default (LGD) for each
    loan
Example: Safe v. Risky Loans

   Safe loans:
     – Over a 1-year period, only 0.25% of these loans
       default
     – If a loan defaults, the bank only loses 1% on the
       outstanding amount
   Risky loans:
     – Over a 1-year period, 1% of loans default every
       year
     – If a loan defaults, the bank loses 10% of the
       outstanding amount
Example: Safe v. Risky Loans
(continued)

   For a $100 million portfolio of the safe loans,
    the bank would expect to see $250,000 in
    defaults in a year and a loss on the defaults
    of $2500
    –   ($100 million X .25% = $250,000)
    –   ($250,000 X 1% loss rate = $2500)
Example: Safe v. Risky Loans
(continued)

   For a $100 million in a risky portfolio the
    bank would expect to see $1 million in
    defaults in a year and a loss on the defaults
    of $100,000
    –   ($100 million X 1% = $1 million)
    –   ($1 million X 10% = $100,000)
Goal of Pillar I

   Although simplistic, this example
    demonstrates what Pillar I is trying to achieve
    –   If the bank’s own internal calculations show that
        they have extremely risky, loss-prone loans that
        generate high internal capital charges, their
        formal risk-based capital charges should also be
        high
    –   Likewise, lower risk loans should carry lower risk-
        based capital charges
Complexity of Pillar I

   Banks have many different asset classes
    each of which may require different treatment
    –   Each asset class needs to be defined and the
        approach to each exposure determined
   Minimum standards must be established for
    rating system design, including testing and
    documentation requirements
    –   The proposals must be tested in the real world
Assessing Basel II

   To determine if the proposed rules are likely
    to yield reasonable risk-based capital
    requirements within and between countries
    for banks with similar portfolios, four
    quantitative impact studies (QIS) have been
    undertaken
Results of Quantitative Impact Studies

   Results of the QIS studies have been
    troubling
    –   Wide swings in risk-based capital requirements
    –   Some individual banks show unreasonably large
        declines in required capital
   As a result, parts of the Accord have been
    revised
Operational Risk

   Pillar I also adds a new capital component
    for operational risk
    –   Operational risk covers the risk of loss due to
        system breakdowns, employee fraud or
        misconduct, errors in models or natural or man-
        made catastrophes, among others
Pillars II and III

   Progress has also been made on Pillars II
    and III
    –   Pillar II focuses on supervisory oversight
    –   Pillar III looks at market discipline and public
        disclosure
Pillar II

   Supervisory Oversight
    –   Requires supervisors to review a bank’s capital
        adequacy assessment process, which may
        indicate a higher capital requirement than Pillar I
        minimums
Pillar III

   Market discipline and public disclosure
    –   The United States is currently in the forefront of
        disclosure of financial data
            SEC disclosure requirements for publicly traded banks
            Bank regulators require quarterly filing of call reports for
             all banks
    –   U.S. authorities are currently considering what
        banks should publicly disclose about their Basel II
        calculations
U.S. Implementation of Basel II

   Based on results for QIS4, which show the
    potential for substantial declines in capital,
    the U.S. banking regulators have proposed a
    revised implementation timeline
    –   The revised timeline includes a minimum three-
        year transition period
Revised U.S. Timeline for Basel II
Implementation


          Year     Transistional Arrangements
   2008          Parallel Run
   2009          95% floor
   2010          90% floor
   2011          85% floor
U.S. Implementation of Basel II
(Continued)

   After 2011, an institution’s primary federal
    supervisor will assess the institution’s
    readiness to operate under Basel II
    –   Institutions will be assessed on a case-by-case
        basis
    –   Further revisions to the floors are anticipated
    –   Both Prompt Corrective Action and leverage
        capital requirements will remain
Basel I-A: The Search for Equal
Capital Treatment

   In the U.S., concerns that Basel II could give
    those banks operating under it a competitive
    advantage over other banks has resulted in a
    proposal called Basel 1-A
   Basel 1-A is designed to modernize the way
    all U.S. banks and thrifts calculate their
    minimum capital requirements
Implications

   The practices in Basel II represent several
    important departures from the traditional
    calculation of bank capital
    –   The very largest banks will operate under a
        system that is different than that used by other
        banks
    –   The implications of this for long-term competition
        between these banks is uncertain, but merits
        further attention
Implications

   Basel II’s proposals rely on banks’ own
    internal risk estimates to set capital
    requirements
    –   This represents a conceptual leap in determining
        adequate regulatory capital
   For regulators, evaluating the integrity of
    bank models will be a significant step beyond
    the traditional supervisory process
Implications

   The proposed Accord will elevate the
    importance of human judgment in the
    process of capital regulation
    –   Despite its quantitative basis, much will depend
        on the judgment of banks in formulating their
        estimates and of supervisors in validating the
        assumptions used by banks in their models
Work Continues

   During the past 3 years the FDIC has
    expressed its concern that the proposed
    Accord will result in banks having too little
    risk-based capital
   Work continues on recalibrating the
    proposals and a workable solution is
    expected
Implications
Additional Data Needed to Counterbalance to Changes in
Environment
                                • Changes in environment
                                  necessitate changes in risk
   Higher
  Leverage
                                  analysis for banks and
                                  supervisors/insurers
                                     • Additional information will be
  Unproven
                   Improved Risk
                                       needed to:
   Rating
                    Management
  Systems                               Inform policy development.
                                        Supplement other sources of
   Evolving          Three Year          risk information used in
   Control         Floors/Leverage       supervisory resource planning
  Structures            Ratio
                                         and overall risk assessments
                                        Serve as an input into deposit
                                         insurance pricing and overall
                                         insurance funds adequacy
                                         analyses
Why XBRL ?

   Internal ratings based and standard approach
    measures require complex data model
    –   Common data requirements flow from Accord and
        Quantitative Impact Studies (QIS I – IV)

    –   Domestic and international comparisons needed to ensure
        consistent application

    –   Taxonomy needed to compare banks’ internal ratings of
        similar and diverse risks
 Common Data Elements Flow from Accord:
 Standardized Internal Risk Estimates

        Data Types                    Reporting Granularity
                 Internal Risk
                   Estimate          Summary
 Exposure                              Data




                 Other
                 LGD
                 EAD
                  PD


 Wholesale       XM  X   X   X   -
   Retail        X   X   X   -   -         Portfolio
                                           Level Data
Securitization   -   -   -   -   X
   Equity        -   -   -   -   X
                                            Individual
 Market Risk     - - - - X              Exposure Data for
 Operational                             All Transactions
                 - - - - X
    Risk
Why XBRL ?
   Internal ratings based measures and standard approach require
    complex data model

    –   Supervisors need detailed information to qualify banks for
        advanced approaches (IRB, AMA, and Market Risk)

    –   Data can be shared across different supervisory regimes

    - Independent of systems, platforms, geography and language
       translation
Consistent data needed to help identify risk estimates
that may be inconsistent with peer estimates.
% of Wholesale Exposures
40
           Bank’s PD Distribution Mapped to S&P Rating Scale

35         Peer Banks’ PD Distribution Mapped to S&P Rating Scale


30


25


20


15


10


 5


 0
     AA or better    A to AA       BBB to A     BB to BBB       B to BB   >B

Follow-up: Can differences between Bank’s PD and benchmark be adequately
explained by differences in risk?
Why XBRL ?

   XBRL provides a framework for complex data
    model
    –   Open standard facilitates reuse and innovation
    –   Analysts can spend more time analyzing data
    –   Reduced reporting burden, especially for
        organizations operating in multiple jurisdictions
Why XBRL ?


   A standard is needed in any case.
Why XBRL ?




        FINIS