CFA Level II Exam · Credit Analysis Models

Reduced-Form Models of Credit Risk for CFA Level II

Updated 7 October 2026 · Fact-checked

Reduced-form models treat default as a random event that arrives at a rate, called the default intensity, set from outside the model. Recovery is also an outside assumption. Default is not linked to a firm's asset value. You solve questions by identifying intensity, recovery and the pricing framework given in the vignette.

Understand Reduced-Form Models of Credit Risk

Start with the problem. Structural models say a firm defaults when its asset value falls below its debt. They need asset value and asset volatility, which you cannot observe directly. Reduced-form models skip that step.

A reduced-form model treats default as a statistical process. The key input is the default intensity (also called the hazard rate), which gives the probability of default over a short period, given that the firm has survived so far. The intensity is not derived from the balance sheet. It is estimated from market data, such as bond spreads or CDS spreads, or from observable variables like macro factors and firm ratios.

Recovery is also an exogenous assumption. The usual one is a fraction of face value, or of market value just before default, that the investor gets back. Loss given default is 1 minus the recovery rate.

Because default is not tied to a firm's asset value, it can come as a surprise. In a structural model, default is predictable as asset value drifts toward the debt barrier. In a reduced-form model, default can occur at any time with no warning. Intensity can also change with the economy, so it can be a function of observable variables.

The trade-off: reduced-form models fit market prices well and are easy to calibrate, and they suit complex or non-traded issuers because they need no asset value. But they give less economic insight into why a firm defaults, and the intensity and recovery are assumptions that can be wrong or unstable. Reduced-form models are therefore used mostly for pricing and for fitting market data.

Key formulas to remember

Approximate default probability over a short period: PD ≈ λ × Δt; λ is the annual default intensity (hazard rate) and Δt is a short time in years. Valid only for small λΔt.
Survival probability with constant intensity: P(survive to t) = e^(−λt); Assumes λ is constant. Default probability to time t is 1 − e^(−λt).
Loss given default: LGD = 1 − Recovery rate; Recovery is an exogenous assumption in reduced-form models.
Approximate credit spread link: Spread ≈ λ × LGD = λ × (1 − R); A rough relationship under risk-neutral measures. Intensity here is the risk-neutral one.
Expected loss: EL = PD × LGD; Use the same time horizon for PD as for the loss you want.

How to solve Reduced-Form Models of Credit Risk questions

Use this approach for any reduced-form question in an item set.

1Read the question first and decide whether it asks you to compare models, identify assumptions, or calculate something.
2Find the exogenous inputs in the vignette: default intensity (hazard rate), recovery rate, and the time horizon.
3If it is a comparison, ask whether default depends on firm asset value. If yes, it is structural. If it is driven by an intensity process, it is reduced-form.
4For a calculation, convert to the right horizon. Use e^(−λt) for survival with constant intensity, or λ × Δt for a short period if the question suggests it.
5Compute loss given default as 1 minus recovery, then expected loss as PD × LGD, or spread ≈ λ × LGD.
6Check whether the intensity is risk-neutral (from market prices) or real-world (from historical data). Do not mix them.
7Check units: annual intensity with years, not months, and percentages converted to decimals.
8Choose the option that matches your result and the reasoning on assumptions or limitations.

Quickest way: Three-check shortcut

When to use it: Use it when you are short on time and the question is conceptual or needs one simple calculation.

Ask: does default depend on asset value? If no, it is reduced-form, and default can be a surprise.
Spot λ and R in the vignette. Compute LGD = 1 − R.
Spread ≈ λ × LGD, or PD over one year ≈ 1 − e^(−λ). Pick the closest option and move on.

Common mistakes in Reduced-Form Models of Credit Risk

Saying reduced-form models link default to the firm's asset value.
Students mix them up with structural models, which are option-based.
Fix: Remember: structural = asset value vs debt. Reduced-form = exogenous intensity, so default can be a surprise.
Treating recovery as something the model derives from asset value.
In structural models, recovery follows from asset value at default.
Fix: In reduced-form models, recovery is an exogenous assumption you are given.
Using the intensity directly as a probability over many years.
Students forget that λ is a rate.
Fix: For multi-year horizons use 1 − e^(−λt) under constant intensity.
Forgetting to multiply by LGD when linking spread to intensity.
Students equate the spread with the hazard rate.
Fix: Spread ≈ λ × (1 − R). Spread equals λ only if recovery is zero.
Claiming reduced-form models explain why a firm defaults.
The model fits prices well, so it feels explanatory.
Fix: State the limitation: little economic insight, and the assumptions on intensity and recovery may be unstable.

Worked examples

Example 1

Vignette: An analyst uses a reduced-form model for a bond issuer. The model assumes a constant annual default intensity of 2.0% (risk-neutral) and a recovery rate of 40% of face value. Q1: What is the approximate credit spread? Q2: What is the probability the issuer survives three years? Options for Q2: A) 94.18% B) 94.00% C) 98.02%.

Show the solution

Q1: LGD = 1 − 0.40 = 0.60.
Spread ≈ λ × LGD = 0.02 × 0.60 = 0.012, or 1.20%.
Q2: Survival = e^(−λt) = e^(−0.02 × 3) = e^(−0.06).
e^(−0.06) ≈ 0.9418, so 94.18%.
Option B (94.00%) is 1 − 3 × 0.02, a linear shortcut that ignores compounding. Option C (98.02%) is e^(−0.02), the one-year survival probability, not the three-year figure.

Answer: Q1: about 1.20%. Q2: A) 94.18%.

Example 2

Vignette: A fund manager compares two approaches for a private firm with no traded equity. Model X ties default to the firm's asset value falling below its debt. Model Y uses an exogenous default intensity linked to macroeconomic variables and an assumed recovery rate. Q1: Which model is reduced-form, and which suits the private firm better? Q2: Which statement about Model Y is correct? A) Default is predictable from asset value. B) Default can come as a surprise. C) Recovery is derived from asset value.

Show the solution

Q1: Model Y uses intensity and an assumed recovery, so it is reduced-form.
A private firm has no traded equity, so asset value and volatility are hard to observe. Model Y needs only intensity inputs, so it suits better.
Q2: In a reduced-form model default is a random event at an intensity, not a barrier crossing, so it can be a surprise.
A and C describe structural features, so they are wrong.

Answer: Q1: Model Y is reduced-form and suits the private firm better. Q2: B.

Exam tips

Expect comparison questions: list what each model needs as input, what drives default, and how recovery is set.
Always spot which inputs are exogenous in the vignette; the question usually hinges on them.
For calculations, note whether the horizon is one year or several, and use e^(−λt) for multi-year survival.
Know the limitations: weak economic insight, assumption-driven, and intensity may not be stable over time.
No marks are lost for guessing, so answer every question even if you are unsure.

Reduced-Form Models of Credit Risk in other exams

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

Reduced-Form Models of Credit Risk: frequently asked questions

What is the main difference between reduced-form and structural models?

Structural models say default occurs when asset value falls below debt, so they need asset value and volatility. Reduced-form models treat default as a random event with an exogenous intensity. They do not use asset value.

What is default intensity?

It is the hazard rate: the rate at which default arrives, given that the firm has survived so far. Over a short period, default probability is about λ × Δt.

Why can default be a surprise in a reduced-form model?

Default is a random event driven by the intensity, not a threshold the asset value crosses. So there is no steady drift toward default that signals it in advance.

What are the limitations of reduced-form models?

They give little insight into why a firm defaults. Results depend on assumed intensity and recovery, which may change over time or differ between market-implied and historical estimates.