Internal Ratings: Leveraging ERM For Regulatory and Business Value Enterprise Risk Management Symposium Concurrent Session: A5 Christopher Whalen Managing Director Institutional Risk Analytics May 3, 2005
Dec 19, 2015
Internal Ratings:Leveraging ERM For
Regulatory and Business ValueEnterprise Risk Management Symposium
Concurrent Session: A5
Christopher Whalen
Managing Director
Institutional Risk Analytics
May 3, 2005
Institutional Risk Analytics
How to Leverage ERM?
• Control your data by structuring internal and external information using industry standards such as XML/XBRL. Get good peer, customer data.
• Control your business by imbedding risk analytics tools and performance metrics into all aspects of management process.
• Control your disclosure by displaying your financial results, performance and risk metrics using same structured data and benchmarks.
Institutional Risk Analytics
ERM Themes
• Mark to Model: For banks, regulators are aligning risk creation and control responsibilities under a portfolio-based discipline that rewards modeling skill, execution and risk avoidance.
• Internal Controls: For all companies, onus is O&Ds to maintain internal systems that track how risks are taken and mitigated, document extraordinary events, and make disclosure timely and accurate.
Institutional Risk Analytics
Interweaving Regulations
Basle II:• New Capital Accord
focuses on credit and operational risk. All performance and risk factors must all be quantified and projected via an internal ratings process that is transparent and documented.
Sarbanes-Oxley:• Regulatory response
to corporate fraud and management failures. Mandates normative roles for O&Ds, auditors and counsel in review and disclosure process to assure adequacy of internal operational and financial controls.
Institutional Risk Analytics
Op-Risk Event Horizon
External Factors
Market/Credit/
Business Risk
OperationalRiskInternal
Factors
Visible to Management
Not Visible toManagement
Institutional Risk Analytics
Risk Segments
• External: Those factors beyond the horizon for management. Utilize macro economic, actuarial and other indicators to monitor/price these invisible risks.
• Internal: Those factors that may be directly observed by management. Utilize internal financial, op-risk metrics to benchmark business and validate controls.
Institutional Risk Analytics
Op-Risk Factors
Market/Credit/
Business Risk
OperationalRisk
• Act of God
• Execution
• Terrorism
• Fraud
• Diligence
• Strategy
• Competitor
• Investment
• Vendor
• Market
• Ethics
InternalFactors
• Governance
• Technology
• Systems
• Procedures
• Counterparty
• Counterparty
Institutional Risk Analytics
Risk Opinion Components
• Absolute Tests– Patterns that indicate elevated vulnerability
to potentially catastrophic events.
• Sudden Motion Tests– Sudden deviations from past behavior or
extreme volatility in behavior.
• Outlier Behavior Tests– Statistically significant deviations from
industry peers.
Institutional Risk Analytics
Indicator Limitations Credit Analysis
Most credit/risk ratings are based on liquidity models that depend on market prices (EMH) that are vulnerable to manipulation by market makers. Quarterly tie-outs to as-reported fundamentals are not mandatory. While useful for momentum investing, Merton models provide little if any forewarning of fraud and/or restatement.
Behavioral Analysis
Behavioral indicators provide visibility into areas where credit ratings do not cover, but also frequently miss events actually detected by Merton models! Low correlations and high false positive rates vs. historical event distributions suggest additional validation needed to achieve a defendable degree of confidence for SOX and/or Basel II compliance.
Institutional Risk Analytics
Modeling Alternatives
• Path 1: Modify Existing Credit Models to Cover Outlier Behavior Risks– Addition of new factors then recalibration of Regression-
Scoring Models.– New modeling to account for “market inefficiency” risk to
correct assumptions of Merton models.– Firm Specific Event Risk Modeling for Cash-Flow
Simulations.
• Path 2: Add a specific behavioral risk profile to capture and score subjects exceeding profile boundary.– Add specific fundamental factors to risk rating profile.– Create/maintain fundamental profile to demonstrate
diligence regarding operational factors.
Institutional Risk Analytics
Ratings System Design
AgencyPolicy
Statements
Academicand
IndustryTreatments
Selection ofTest MethodsAnd Metrics
LegalFilingsSourceData
Numerical Benchmarks
Temporal Trends
Peer ComparisonsObjective: select a test set that covers the relevant risks.
Institutional Risk Analytics
The Ratings Game
• To validate any risk management system, must benchmark projections against actual.
• Advantage of a common approach to risk taking and risk management is that business performance becomes expression of both.
• By examining financial performance in detail, can assemble a profile of business and operational factors.
Institutional Risk Analytics
Ratings: Basel II
Institutional Risk Analytics
Basel II Bank Metrics
• Business Performance:– Profitability, productivity, solvency; actual vs.
projected vs. peers.
• Credit Risk Factors:– P(D), LGD, M, EAD in aggregate and by key lending
classes; actual vs. projected vs. peers
• Operational Risk Exposure:– Management decisions and external factors
generating out of “norm” events; actual vs. projected.
Institutional Risk Analytics
Summary Risk Profiles
(December 2004)Name Assets ROA ROE LGD P(D) WAM EAD RatingJPM $967 0.20% 3.23% 75.1 62.4 3.54 69.7 BB
Citibank $964 1.50% 18.80% 77.4 182.6 1.68 160.5 BBBOA $771 1.50% 21.20% 57.7 30.6 7.83 64.23 BBB
WACH $390 1.30% 14.10% 52.7 22.2 5.59 82.32 BBBSource: Federal Deposit Insurance Corp/Institutional Risk Analytics
Basel II: A qualifying IRB rating system must have two separate and distinct dimensions: (i) the risk of borrower default, and (ii) transaction-specific factors.
Institutional Risk Analytics
Basel II Risk Modeling
ActualValue
IRB Model Predictionfalls within target range.
Performance“Norm”ToleranceLimit
Excessive deviation triggers PCA to recalibrate IRB modeling system.
Institutional Risk Analytics
Basel II Credit Metrics
Layer 1
Revised Standardized
Approach
• Relies on external ratings based on borrower category.
• Provisions for added risk sensitivity weight factors.
Layer 2
Foundation IRB
Layer 3
Advanced IRB
Probability of Default (PD)
Internal Bank Estimate
Internal Bank Estimate
Loss Given Default (LGD)
Supervised Factor Internal Bank Estimate
Maturity (M) Supervised Factor Risk Weight Adjustable
Exposure at Default (EAD)
Supervised Factor Internal Bank Estimate
Institutional Risk Analytics
Mark to Actual
• Use enhanced internal metrics to set up a closed loop assurance function to ensure that bank IRB systems that rely on forward projection statistics are marked to actual.
• Control loop can be used both internally by risk officers and auditors, and externally by regulators, to ensure a healthy system of checks and balances.
Institutional Risk Analytics
Macro Implications
• Basel II credit risk reporting factors will become a de facto standard for comparing all banks among credit analysts and buy-side investors.
• Basel factors will be used to streamline regulation, explain operating policy, and facilitate peer comparisons. Banks will rate themselves in real time vs. Basel projections.
• Periodic reporting will eventually move to monthly frequency. Internal systems will “mark to actual” vs. real time call report.
Institutional Risk Analytics
Basel II: M&A EffectLarge Bank Peer Group -- LGD
40.00
50.00
60.00
70.00
80.00
90.00
100.00
1999
03
1999
06
1999
09
1999
12
2000
03
2000
06
2000
09
2000
12
2001
03
2001
06
2001
09
2001
12
2002
03
2002
06
2002
09
2002
12
2003
03
2003
06
2003
09
2003
12
2004
03
2004
06
2004
09
2004
12
JPMCitibankBOAPeer Avg
Institutional Risk Analytics
Macro Implications
• Bank risk estimation methods will be benchmarked and certified against legal filings of periodic regulatory reports, providing a very clear and very public measure of performance.
• Prompt corrective action (PCA) to eliminate excessive deviations between models and filings will become a key regulation mechanism.
• Institutions that excel at execution and managing the rate of external events will win the competitive race.
Institutional Risk Analytics
US Bank National AssociationP(D) (cumulative by year through 12/04)
0
20
40
60
80
100
120
140
160
180
1999-03 1999-12 2000-09 2001-06 2002-03 2002-12 2003-09 2004-06
Bank Unit
Peer Avg
Basel II: P(D) Profile
Institutional Risk Analytics
Ratings: C&I Profile
Institutional Risk Analytics
Selected C&I Metrics
• Business Performance:– Profitability, productivity, solvency, actual vs.
projections, and vs. peers.
• Risk Factors:– Behavioral analysis, earnings & assets quality, long-
term business and competitive trends.
• Operational Risk:– Quality of management decisions and external factors
generating out of “norm” events, audit, governance.
Institutional Risk Analytics
Summary Risk Profiles
C&I Rating Factors12/31/2004 Solvency Debt Service Gross Margin Sales Growth Z-Score RatingF n/a n/a Downturn Unstable Danger BBB-GM n/a n/a Stable Stable n/a BBB-DCX Solvent Marginal Stable Stable Danger BBB+TM* Solvent Adequate Stable Stable Warn AAA* 3/31/04
Source: IRA Corporate Monitor/CoreData/S&P
Based on the assessment of these and other factors, the risk manager may then assemble an independent rating for the subject companies.
Institutional Risk Analytics
C&I Profile
Visteon Corporation(millions)
Source: IRA Corporate Monitor/CoreData
2004 2003 2002 2001 2000Net Income -1,489 -1,213 -352 -118 270EBITDA -438 -454 550 548 1,337Solvency SOLVENT SOLVENT SOLVENT SOLVENT SOLVENTDebt Service MARGINAL MARGINAL ADEQUATE ADEQUATE ADEQUATEAsset Qual DECLINING STABLE STABLE STABLE STABLEZ-Score N/A WARN OK OK OK
Based on the assessment of these and other factors, the risk manager may then assemble an independent rating for the subject companies.
Institutional Risk Analytics
Analytical Challenges
• There are 20 million companies in the US, but only 15,000 are public.
• Credit officers frequently rely on “mutual revisions” rather than maintain informed, independent credit opinions, especial on “in-betweens.”
• Must extend range of default prediction tools from current 20% probability to 50%, in particular to detect fraud, restatement.
Institutional Risk Analytics
The “In-Betweens”
Institutional Risk Analytics
ERM Challenges
• Structuring the collection and warehousing of information including the migration and sweetening of existing historical data.
• Achieving Advanced IRB compliant end-to-end risk analysis systems deployed at enterprise level by the 2007 date for Basle II start.
• Workable expense containment strategies that increase rate of data throughput and analysis, but lower “per transaction” cost.
Institutional Risk Analytics
Who is Institutional Risk Analytics?
• IRA is part of a movement within the financial analytics community to broaden risk measurement tools to include behavioral elements described by fundamental factors.
• IRA builds customized risk systems for processing public & privileged data, and publishes research on companies and topics affecting financial policy & regulation.
Institutional Risk Analytics
Contact Information
Corporate Offices
Lord, Whalen LLCdba Institutional Risk Analytics14352 Yukon AvenueHawthorne, California 90250Tel. 310.676.3300Fax. [email protected]
WEBSITE:www.institutionalriskanalytics.com
For inquiries contact,
R. Christopher Whalen
Managing Director
Head of Sales and MarketingTel. 914.827.9272
Fax. 914.206.4238
Cell. 914.645.5304