Credit Portfolio Models

Question 1

Why are credit VaR models needed for loans but not for general portfolio returns?

Answer 1

We cannot assume a normal distribution of loan returns, as they have limited earnings which are highly likely, and a limited probability of huge losses.

Question 2

What are credit events in the creditmetrics approach?

Answer 2

Deafault is just one among a set of credit events.
Credit events = all events that can alter the value of a loan (defaulut is only the most significant one)
But all changes in the borrower’s rating are credit events

Question 3

What are five major types of credit events?

Answer 3

1. Bankruptcy: legal process that ensues when an individual or organization is unable to repay their outstanding debts.
2. Obligation acceleration: Debt contracts often have covenants that allow the creditor to demand their outstanding loans immediately.
3. Payment default: individual or organization unable to make payments on their debts on a timely basis.
4. Repudiation/Moratorium: when a governmental authority, either refuses to recognise or challenges the validity of one or more obligation of the reference entity, or imposes a moratorium thereby postponing payments on one or more of the obligation of the referece entity
5. Debt restructuring: change in the terms of the debt, causing the debt to be less favorable to debt holders

Question 4

What defines a company rated A on the scale by S&P.

Answer 4

If the bank’s raters know how to do their job, it is unlikely that an A-rated borrower will fail, or migrate to calss C, within one year.

Question 5

Why are ratings important for credit portfolio models?

Answer 5

The different ratings pay different risk-adjusted rates (they still depend on the bond’s LGD)

Question 6

What is the formula for calculating forward rates from spot rates?

Answer 6

₁r₁ = (1+r₂)²/(1+r₁) - 1

Question 7

What are the forward rates useful for in credit portfolio models?

Answer 7

We can calculate value of the loan for different ratings after one year, based on the spot rate associated with each rating, and the value at default (which depends on the recovery rate). By taking the weighted average of the value by ratings + default with their respectivee probabilities we can calculate the μ (expected value of the loan)

Question 8

How do we calculate the VaR of a portfolio with multiple loans that are correlated?

Answer 8

We take their possible values based on the possible future ratings, we arrange the values from lowest to highest and search for the accumulated probability cutoff which coincides with the desired confidence level. We then take the difference between the expected value and the value associated with the desired percentile.

Question 9

What are advantages of the creditmetrics approach to credit portfolio models?

Answer 9

1. Uses objective and forward looking market data.
2. Evaluates the portfolio market value.
3. Takes into account migration risk.

Question 10

What is migration risk?

Answer 10

The risk taht a bond will move to a lower credit rating in the future.