Chapter 23: Assessment of credit risk Flashcards
Places to obtain information to assess credit risk RCPP
- Rating agencies
- Counter party itself
- Publicly available data
- Proprietary data
Challenges to model credit risk DECS
- Data shortage
- Lack of extreme events
- Correlation of defaults
- Skewness of the distribution
Factors on which qualitative assessment is based FINES C
- Financial ratios
- In-person meetings with counterparty
- Nature of the borrower – industry, assess jockey
- Economic indicators
- Security on the loan
- Contractual obligation
Merits of quantitative models WD MERC
- Wide range of factors can be considered
- Dynamic – can be easily changed
- Meaning of assessments can change
- Excessive subjectivity may be present
- Responsiveness of ratings may be low – quick changes due to change in environment or circumstances of counterparty
- Consistency is difficult to achieve
Flawed assumptions of Merton model and advantages of the KMV model F BOARDS LOB
o Frictionless markets
o Bond is zero coupon – a single default instance
o Observable traded security – asset price
o Asset price follows geometric Brownian motion
o Risk free rate is deterministic and constant
o Default results in liquidation
o Sentiment/market’s effect on asset price not considered
o Liability structures that are more complex can be considerd
o Observability of assets X0 not assumed
o Bonds which are coupon paying can be assumed
Loss on default depends on LIERS C
- Legal jurisdiction
- Industry
- Economic cycle
- Rights of other creditors
- Seniority of the debt
- Collateral – how marketable is it?
Merits of reduced form models to model credit risk SPUD CLAUD
- Stability in results
- Public information not required
- Unintuitive time-homogeneity assumption
- Default probabilities are assumed to be known
- Credit rating assumed to capture PD fully
- Lack of credit rating for some companies
- Agencies’ ratings of companies may differ
- Unavailable ratings
- Distinct groups of ratings are few – not very granular
Credit portfolio models MESS
Migration models MV
Econometric/actuarial models
Survival models - Time to default
Structural models MV
Properties of the Merton model VODAS
- Volatility of equity
- Option pricing theory: Shareholders have call option on company assets (Black Scholes)
- Debt value derived: Zero coupon bond
- Asset value derived: Debt + Equity
- Strike price of option = nominal value of debt
Features of Credit migration model TAP
- Transition probability matrix – Markov Chains
- Apply to current credit rating
- PD calculated - migration to default
Multivariate structural models KAMC
- KMV/Merton models can be constructed to model asset values organisations in portfolio
- Asset values of counterparties in the portfolio modelled
- Multivariate models (MVN, MVT) used to model logarithm of asset values and interactions between companies
- Copulas can be used to allow for interactions
Elements of a company modelled with CreditMetrics BARE
- Bond value
- Asset value
- Rating
- Equity value
Assumptions of the credit metrics approach PRICL
- PD assigned to each credit rating
- Rating is a function of volatility and value of assets
- Indices of country and firm volatility used
- Correlation of assets assumed the same as equity
- Log-normal behavior of assets
Features of Survival models TECHS
- Time of default relationship in portfolio of bonds described
- Exponential function for constant hazard rate
- Copulas used in the description
- Hazard function used to describe survival CDF
- Several ways to calculate hazard rate (Merton, credit rating, default history, economics)
