Credit Risk

Credit Risk

1. Rating Systems

   • Risk quality covers both the probability of default and the recoveries in the event of default
   • A summary indicator of the risk inherent in an individual credit
   • Internal Ratings
     • In-house rating System, customisable to the bank's needs.
     • Grades assgined to borrowers or facilities ranking their relative risk
     • Most bank clients are not rated by external credit rating agencies but are rated internally
     • Developed internal risk models for purpose such as
       • Approving loans
       • Portfolio Monitoring and Management Reporting
       • Analysis of adequacy loan loss reserve or capital
       • Profitability and loan pricing analysis
     • Assigned by bank personnel and are not revealed to outsiders
       • External ratings are considered when assigning internal ratings
         • Banks with judgement based system rely heavily on these ratings
           • External rating either dominate or serve as a starting point to Internal rating
     • Components of internal Rating
       • Borrower's risk
       • the role of a supporting entity
         • Combine assessment of that support with the borrower's intrinsic ratings provide the overall rating
           • Formal support
             • Government owned companies
           • Informal support
             • Holding company and subsidiary
       • Facility (Transaction) Risk
     • 3 Main categories identified by Models Task Force of Basel Committee on Banking supervision
       • Statistical-based processes
         • A default probability model or other quantitative model is essentially the sole basis for determining a rating for exposure within certain portfolios.
         • Include both quantitative (financial ratios) and some qualitative but standardised (industry, payment history) factors.
         • Eg. Use of credit scoring and the KMV approach
         • Prominent role in small corporate lending than medium/large corporates
       • Constrained expert Judgement-based processes
         • Ratings are based primarily on statistical default/scoring models
         • Allow those assigning a rating to adjust it to an explicitly limited degree based on judgement factors.
         • Quantitative tool determines the grade, but the raters may adjust the final grade up or down based on their judgement
         • Quantitative and judgement fators are explicitly assigned a weight in the final rating, thereby effectively limiting the influence of judgemental considerations
       • Expert judgement-based processes
         • Most extensively used type of internal rating process in which ratings are assigned using considerable judgemental elements, where the relative importance given to each of them is not formally constrained.
         • Statistical models may still have a role to play
         • The rater has the discretion to significantly deviate from statistical model indication in assigning a grade
   • External ratings
     • To evaluate and quantify credit risks, within a context of effective benchmarking of risks across industries and countries
     • Aim is to provide lenders and investors with independent and objective credit opinions
     • 2 Types of rating agencies
       • International
         • E.g Standard and Poor, Moody's Investor Service, Fitch Group
         • BBB- and above are investment grade
       • Local
         • Do not address direct sovereign risks, such as the possible imposition of foreign exchange controls, and only address country risk factors among different sectors
         • Cannot be compared with between countries and on international scale
         • Advantage in providing a more precise ranking of relative credit risk within a county.
         • Eg. ICRA in india, Korea investors Service
     • Applies to
       • Sovereign, state and municipal governments
       • Financial Institutions
       • Corporates
     • Applies for
       • Debt Issues
       • Preferred stock
       • Bank Loans
       • Structure Financing
     • Exist only for large listed companies
       • Statistical analysis of default can be misleading because the sample of borrowrer rated are not representative of banks' portfolio
     • Criteria for assessing
       • Country risk, industry risk, market position, management quality, profitability, capital structure and financial ratios.
     • Rating Process
       • Visit to the company, meet with the management and discuss company policy
       • Some uses public information only.

2. Constituents of credit risk

   • Default Risk
     • The adequate quantitative measure of default risk is the probability of a default occurring
     • Default such as missing a payment obligation, breaking a covenant, entering a legal procedure or economic default (When economic value < value of outstanding debt)
     • Default has occurred when a contractual payment has been missed for @ least 3 months
     • Measure of POD are either credit ratings or historical statistics on defaults
     • Higher the rating, the lower the cost of borrowing
   • Exposure Risk
     • The amount at risk in the event of default excluding recoveries.
     • Risk is generated by the uncertainty regarding future amounts at risk.
       • Type of commitment given to the borrower sets an upper limit on possible future exposures
       • Credit line with repayment schedule, exposure risk can be considered small
       • Borrower may draw on these lines of credit within a limit set by the bank, as borrowing needs arise.
   • Recovery Risk
     • Recoveries in the event of default are unpredictable, depending on typer of default and guarantees received
     • Requires legal procedures, expenses and significant lapse of time
     • Recoveries through collateral, guarantees and covenants
       • Collateral
         • Reduces credit risk - if taken over and sold at significant value
         • In the form of cash, financial assets of fixed assets
         • May not be able to sell @ market price
         • Mitigate credit risk into recovery risk + asset value risk
       • Guarantees
         • Contingencies given by 3rd Parties
           • Parent co. to honour the obligation in the event of default
             • Consideration of joint probability of default by both guarantor and the borrower
     • In worst cause, there may be no recoveries because company is sold or liquidated
   • Expected Loss
     • The product of the loss given default and the probability of default
       • L = D x X x (1-R)
         • L = Expected Loss, D = Given default can be viewed as the product of a random variable characterising default, X = exposure, and R = uncertain recovery rate
   • Migration risk

3. Credit risk Management

   • Relied on a number of contractual mechanisms to control credit risk of lending
     • 1. Requiring higher interest rate spreads and fees on loans to more risk borrowers
       • 2. Restricting or rationing loans to more risk borrowers
         • 3. Requiring enhanced seniority for the bank over the assets of risk borrowers
           • 4. Diversifying across different types of risky borrowers
             • 5. Placing more restrictive covenants on risk borrower's actions
   • Credit allocation decision
     • The follow up of credit commitments, monitoring and reporting processes
     • Managing credit risk by reducing unsystematic risk through loan portfolio diversification
     • Employed credit derivatives to hedge credit risk
   • Credit enhancement
     • Seeks to reduce the amount of loss in the event of default through increasing the recovery rate
     • Uses covenants and guarantees
       • Covenants are designed to aid the active monitoring of risks and thus trigger pre-emptive actions when a borrower's credit standing deteriorates
         • Creates incentive for the borrower to comply with specified conditions
       • Aim is to initiate negotiations before default rather than to cause default
   • Loan sales
     • Occurs when a bank originates a loan and then sells it either with or without recourse to an outside buyer
     • If w/o recourse, it is removed from the balance sheet and bank has no explicit liabilities if loan becomes bad. Buyer bears all the credit risk
     • With recourse, the buyer can put the loan back to the selling bank and thus the bank retains a contingent credit risk liability.

4. Credit risk models

   • Focus on determining the default probability based on the observable attributes of borrowers
     • Ability to measure POD is influenced by the level of info
       • Retail
         • Collect information internally or from specialist credit agencies
       • Large corporate borrowers
         • Publicly available info.
   • Qualitative models
     • If limited public info, banks can gather info of borrower's credit and deposit accounts, or purchase information from external credit agencies.
     • factors influencing can be categorised into
       • Borrower-specific (unique to individual borrower), such as reputation, leverage, volatility of earnings and collateral
       • Market-specific (impact on all borrowers), such as business cycle and level of interest rates
       • rely heavily on qualitative analysis, hence, subjective judgement of bank employees
   • Credit Scoring models
     • Uses the borrower characteristics to calculate a score
       • Represents the applicant's POD or default risk class
     • @ retail level
       • factors include income, age, assets, occupation and location
       • Applies well to retail due to large volume of data
     • Commercial level
       • Factors are cash flow info, financial ratios.
     • Techniques of discrimination
     • Statistical techniques used to quantify
       • Linear probability models
         • Uses past data as inputs to a model to explain the repayment experience on pre-existing loans.
           • Used to forecast repayment probabilities on new loans
       • Logit model
         • Represents an adjustment to the linear probability model to ensure that the estimated range of default probabilities lies between 0 and 1.
       • Linear discriminant models
         • Classification of borrowers into high or low default risk
         • Problem: Discriminate only between 2 extreme cases, default or no default up to a given horizon
     • Failure
       • Type 1: accepting bad credits - resulting in full loss of P + I if materialised
       • Type 2: Rejecting good credits - result in opportunity costs relating to miss profitability opportunities.
     • Success
       • Credit card applications
   • Term structure models
     • Market based method of assessing credit risk exposure and default probabilities
       • Analyse the risk premiums inherent in the current structure of yields on corporate debt to similar risk-rated borrowers
         • Differing spread = credit risk exposures
   • Option-based models
     • Merton (1974)
       • When firm raise funds, it has option to default of repay
         • Due to Limited liabilities for equity holders, borrower's loss is limited on the downside by the amount equity invested in the firm
         • If investment projects are successful, borrower keeps most of the upside returns, after P + i have been paid
         • Characteristics of limited downside with upside potential = holding an option
           • Payoff function for equity holders resembles that of holding a call option
           • Payoff function for debt holders resembles that of selling a put option, since the max amount they receive is the promised payments
       • Key insights is that lender should adjust the required risk premium as borrower's leverage and asset risk changes.
         • enables the estimation of risk premiums and default probabilities
           • Implies borrower's market value of assets and its asset risk are key variables on which to focus in credit risk evaluation
             • Problem: market value of a firm's assets and the volatility of firm's assets are not directly observable
               • Commercial users - Moody's KMV - recognises the problem
                 • Approach relied exclusively on equity market info.
                   • Treats the value of equity in firm as equivalent to holding a call option on the assets of the firm
                     • Proceed to use the stock market value of the firm's shares and volatility of the value of the firm's shares to derive the implied MV of the firm's assets and implied assets volatility.
                       • Calculate the current debt obligations
                         • Resulting expected default frequency (EDF) reflects the probability that the MV of the firm's assets will fall below the promised repayments on its short term debt liabilities within one year.
                           • If value of firm's assets falls below its debt liabilities, it can be viewed as economically insolvent