Skip to content

FRM Part II · FRM Exam Part II

Introduction to Credit Risk Modeling and Assessment

Credit risk modeling estimates how likely a borrower is to default, how much you lose if it does, and how large the exposure is. You combine PD, LGD and EAD into expected loss, then measure unexpected loss and credit VaR. Structural, reduced-form and scoring models supply the inputs.

What this chapter covers

This chapter builds the vocabulary and toolkit for all credit risk work in FRM Part II. You start with the three building blocks: probability of default (PD), loss given default (LGD) and exposure at default (EAD). Together they give expected loss. Then you move to the spread of losses around that average, which is unexpected loss and credit VaR.

The second half covers how PD is estimated. Ratings and transition matrices give a historical view. The Merton model and KMV treat equity as an option on firm assets and give a market-based PD. Reduced-form models treat default as a random event driven by a hazard rate. Credit scoring uses borrower data and statistics to rank risk.

This chapter connects to the rest of the Credit Risk topic, such as portfolio credit risk, counterparty risk, structured credit and capital rules. It also supports Market Risk and Liquidity questions that use spreads and default probabilities. If these basics are weak, later chapters become harder.

Credit Risk Measurement and Management is one of the six topics in the 80-question exam, and this chapter supplies the base for the rest of it. Questions are often applied: you are given figures, you pick the right measure, calculate it and interpret it. The formulas are short, so well-prepared candidates can pick up marks quickly. The traps are in definitions and conditions, such as risk-neutral versus real-world PD, or expected versus unexpected loss. Time spent here also pays off in later credit chapters.

Introduction to Credit Risk Modeling and Assessment: topics in the order to study them

  1. 1Credit Risk Components: PD, LGD, EAD and Expected LossEvery other topic in the chapter uses these terms, so learn them and EL = PD × LGD × EAD first.
  2. 2Unexpected Loss and Credit VaRIt extends expected loss to the loss distribution and capital, so it follows straight after the components.
  3. 3Credit Ratings and Transition MatricesRatings are the simplest source of PD estimates, so you see where inputs come from before the harder models.
  4. 4Structural Models: Merton Model and KMVIt links PD to firm value, debt and asset volatility, and needs the option logic you already know.
  5. 5Reduced-Form Models and Hazard RatesIt is easier to compare with structural models once you have seen them, as it models default timing rather than firm value.
  6. 6Credit Scoring and Quantitative Assessment of BorrowersIt covers statistical tools for retail and corporate borrowers and is best revised last, with all PD approaches in mind.

How to prepare Introduction to Credit Risk Modeling and Assessment

Aim to be able to define, calculate and interpret each measure. Work in short sessions that suit phone study, then test yourself with numbers.

  1. Write the core formulas on one page: EL = PD × LGD × EAD, survival probability from a hazard rate, and the link between spread and PD and LGD.
  2. Practise expected loss with several sets of inputs until you can do it without notes. Always check whether PD is annual or multi-year.
  3. Learn unexpected loss and credit VaR as a gap between a high percentile loss and expected loss. Be clear on what is being measured.
  4. Practise reading a transition matrix: one-year default column, multi-year probabilities by matrix multiplication, and the idea that default is absorbing.
  5. For Merton, remember equity as a call option on assets and risky debt as risk-free debt minus a put option on firm assets, with the strike equal to the face value of the debt. N(−d2) gives the risk-neutral PD; the real-world PD uses the asset drift in place of the risk-free rate. Practise the direction of every effect: higher volatility, higher leverage, longer horizon.
  6. Compare structural, reduced-form and scoring models in a short table of your own: inputs, strengths, weaknesses, and real-world versus risk-neutral PD.
  7. Finish with timed mixed questions, then review each wrong answer and note whether the cause was a definition, a formula or arithmetic.

Common mistakes in Introduction to Credit Risk Modeling and Assessment

  • Mixing up expected loss and unexpected loss

    Fix: Expected loss is the mean. Unexpected loss is the variation around it. Capital is linked to unexpected loss.

  • Using a one-year PD for a multi-year horizon

    Fix: Always match PD to the horizon. Use survival probabilities or matrix powers when the horizon is longer than a year.

  • Treating LGD and recovery rate as the same number

    Fix: Compute LGD = 1 − recovery rate, and check whether recovery is quoted per unit of exposure or of face value.

  • Getting the direction of Merton effects wrong

    Fix: Think of equity as a call. More asset volatility raises equity value but also raises PD. Higher debt or lower asset value moves the firm nearer to default. Remember that N(−d2) is the risk-neutral PD; the real-world PD uses the asset drift.

  • Confusing risk-neutral and real-world default probabilities

    Fix: Spread-implied PDs include risk premia and are usually higher. Use historical data for real-world loss estimates, and market prices for pricing.

  • Treating reduced-form and structural models as interchangeable

    Fix: Structural models link default to firm value and capital structure. Reduced-form models treat default as an exogenous random event (a surprise) driven by a hazard rate, or intensity, often calibrated to market spreads.

Last-day revision: Introduction to Credit Risk Modeling and Assessment

  • Expected loss = PD × LGD × EAD; it is the average loss and is a cost of doing business.
  • Unexpected loss is the variability of loss around expected loss; capital covers it.
  • Credit VaR = loss at a chosen confidence level minus expected loss, under the common definition. Check the question's wording.
  • LGD = 1 − recovery rate, when recovery is stated as a share of exposure.
  • EAD includes drawn amounts plus an expected share of undrawn commitments.
  • Transition matrices show the chance of moving between ratings over a period; default is an absorbing state.
  • Merton: equity is a call option on firm assets, with strike equal to the debt face value.
  • In Merton, higher asset volatility or leverage raises the PD. N(−d2) is the risk-neutral PD; the real-world PD uses the asset drift instead of the risk-free rate.
  • KMV uses distance to default and maps it to an expected default frequency from historical data.
  • Reduced-form models: with constant hazard rate λ, survival probability to time t is e^(−λt).
  • Approximate credit spread ≈ PD × LGD, with a risk-neutral PD that is usually higher than the real-world PD because of risk premia.
  • Scoring models rank borrowers; check discrimination and calibration separately.

Introduction to Credit Risk Modeling and Assessment practice questions

Introduction to Credit Risk Modeling and Assessment in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Introduction to Credit Risk Modeling and Assessment: frequently asked questions

What are the three components of credit risk?

They are probability of default, loss given default and exposure at default. Multiplying them gives expected loss for a single exposure. Exam questions often give two and ask you to find the third or the resulting loss.

Do I need to memorise the full Merton model formulas?

Know the structure well: equity as a call option on firm assets, and the link from distance to default to PD. Be ready to interpret the inputs and the effect of changes. Check the readings for any calculation you may need to perform.

How is credit VaR different from market VaR?

Credit VaR looks at losses from default or rating changes over a long horizon, often one year. The distribution is skewed, with a long loss tail, so a normal assumption is usually poor. Market VaR usually has a short horizon.

How should I use transition matrices in the exam?

Read the row for the starting rating and the column for the ending rating. For multi-year probabilities you multiply the matrix by itself, assuming the process is Markov. Remember that default is absorbing.