FRM Exam Part II · Introduction to Credit Risk Modeling and Assessment

Reduced-Form Models and Hazard Rates Explained

Updated 11 October 2026 · Fact-checked

A reduced-form model treats default as a surprise event that arrives randomly at a rate called the hazard rate (intensity), λ. With constant λ, survival to time t is e^(−λt). Under risk-neutral pricing, spread ≈ λ × (1 − recovery rate), so you can read default intensity from market spreads.

Understand Reduced-Form Models and Hazard Rates

Structural models, like Merton, explain default through the firm's asset value falling below its debt. Reduced-form models skip that. They do not model the balance sheet. They say default is a random event that can occur at any moment, and they model only how likely it is.

The key input is the hazard rate (also called default intensity), λ. It is the instantaneous conditional probability of default: over a very short interval dt, the chance of default, given survival so far, is about λ × dt. If λ is constant, default time follows an exponential distribution. Survival probability to time t is e^(−λt), and cumulative default probability is 1 − e^(−λt).

The hazard rate can be constant, time-varying (a term structure of hazard rates) or stochastic, driven by factors such as the economy. A Poisson process is the usual building block: default is the first jump of the process. Because default arrives as a surprise, short-term credit spreads do not fall to zero for high-quality names, which structural models struggle to explain.

The link to markets is simple. A bond holder loses the fraction LGD = 1 − R if default happens. Under risk-neutral pricing, the credit spread compensates for expected loss: spread ≈ λQ × (1 − R). Here λQ is the risk-neutral hazard rate, implied from prices. It is usually higher than the real-world (physical) hazard rate λP estimated from historical defaults. The gap reflects risk premia: investors demand extra return for default clustering, uncertainty and illiquidity. So never use a spread-implied PD as a forecast of actual defaults. Use it for pricing and valuation, such as CDS and CVA. Use real-world PD for loss forecasting and capital.

Key formulas to remember

Survival probability (constant hazard): Q(τ > t) = e^(−λt); τ is the default time. λ is a constant annual rate. Use the same time unit for λ and t.
Cumulative default probability: PD(t) = 1 − e^(−λt); For small λt, PD(t) ≈ λt.
Time-varying hazard rate: Q(τ > t) = exp(−∫0→t λ(s) ds); With piecewise-constant λ, add λ × length across each interval inside the exponent.
Credit spread and hazard rate: s ≈ λ × (1 − R); Approximation under risk-neutral pricing with a flat curve and continuous compounding. λ = s ÷ (1 − R). R is the recovery rate.
Average (annualised) hazard from cumulative PD: λ = −ln(1 − PD(t)) ÷ t; Use this to convert a cumulative default probability into a constant annual intensity.
Conditional default probability in interval: P(default in (t, t+dt) | survival to t) ≈ λ(t) × dt; Defines the hazard rate. It is a conditional measure, not unconditional.
Risk-neutral vs real-world: λQ ≥ λP in practice; A rule of thumb from observed markets, not a mathematical law. The difference reflects risk premia.

How to solve Reduced-Form Models and Hazard Rates questions

Most questions give you a spread, a recovery rate or a default probability and ask you to convert between them or interpret the result. Work in this order.

1Identify what is given: spread, hazard rate, cumulative PD, or survival probability. Note the time horizon and the recovery rate.
2Decide whether the measure is risk-neutral (implied from prices or spreads) or real-world (from historical defaults or ratings).
3Convert spread to hazard rate with λ = s ÷ (1 − R). Convert units: 150 bps = 0.015.
4Convert hazard to probability: survival = e^(−λt), default = 1 − e^(−λt). Check that t matches the unit of λ.
5For a term structure, add the hazard rates over each interval before taking the exponent.
6State the interpretation: which measure it is, and what it should be used for (pricing vs forecasting).
7Sanity check: λ should be positive and small, PD should lie between 0 and 1, and a higher R means a higher λ for the same spread.

Quickest way: Spread-to-PD shortcut

When to use it: Use when the question gives a spread and recovery rate and asks for an approximate default probability, and options are well separated.

Compute λ = spread ÷ (1 − R).
For short horizons or small λ, PD ≈ λ × t.
Check which options are close. Use the exact 1 − e^(−λt) only if two options are near each other.
If asked which is larger, risk-neutral or real-world, pick risk-neutral unless the question says otherwise.

Common mistakes in Reduced-Form Models and Hazard Rates

Treating the hazard rate as the unconditional annual default probability.
Both are small numbers and both look like annual probabilities.
Fix: Remember that λ is conditional on survival and is an intensity. Convert with 1 − e^(−λt). For small λ the two are close, but they are not the same thing.
Using spread = λ without dividing by (1 − R).
Students forget that only the loss given default is compensated.
Fix: Always write λ = s ÷ (1 − R). With R = 40%, λ is the spread divided by 0.6.
Using a spread-implied PD as a forecast of real defaults.
The numbers look like probabilities, so they are read as predictions.
Fix: Label them risk-neutral. They include risk premia and liquidity effects, so they usually overstate real-world default frequency. Use them for pricing, not loss forecasting.
Mixing time units, for example using a 5-year cumulative PD as an annual hazard.
Students skip the unit check under time pressure.
Fix: Write t next to λ. To get an annual λ from a cumulative PD, use λ = −ln(1 − PD) ÷ t.
Confusing reduced-form with structural models.
Both produce default probabilities and both appear in the same chapter.
Fix: Structural: default when assets fall below debt, so it is predictable in principle. Reduced-form: default is an unpredictable jump driven by an exogenous intensity, calibrated to market prices.
Subtracting hazard rates instead of adding them across periods.
Students mix up survival and default probabilities.
Fix: Cumulative hazard adds across intervals. Survival is the exponential of the negative total.

Worked examples

Example 1

A 5-year USD bond issued by a corporate trades at a credit spread of 180 bps over the risk-free rate. Assume a recovery rate of 40% and a constant risk-neutral hazard rate. Estimate the risk-neutral probability of default within 5 years.

Show the solution

Spread s = 180 bps = 0.018. R = 0.40, so 1 − R = 0.60.
λ = s ÷ (1 − R) = 0.018 ÷ 0.60 = 0.03, or 3% per year.
Survival to 5 years = e^(−0.03 × 5) = e^(−0.15).
e^(−0.15) ≈ 0.8607.
Cumulative default probability = 1 − 0.8607 = 0.1393.

Answer: About 13.9% risk-neutral probability of default within 5 years (λ = 3% a year). This is a pricing measure, not a forecast of real defaults.

Example 2

A bank estimates from historical data that a borrower has a real-world hazard rate of 2% a year. The borrower's 3-year CDS spread is 150 bps and the assumed recovery rate is 40%. Compare the risk-neutral and real-world 3-year default probabilities and explain the difference.

Show the solution

Risk-neutral λQ = 0.015 ÷ 0.60 = 0.025, or 2.5% a year.
Risk-neutral PD(3) = 1 − e^(−0.025 × 3) = 1 − e^(−0.075).
e^(−0.075) ≈ 0.9277, so PD(3) ≈ 0.0723.
Real-world PD(3) = 1 − e^(−0.02 × 3) = 1 − e^(−0.06).
e^(−0.06) ≈ 0.9418, so PD(3) ≈ 0.0582.
Compare: 7.23% against 5.82%. The risk-neutral probability is higher.
Reason: the market spread includes a premium for bearing default risk, such as default clustering, uncertainty about recovery and liquidity, on top of expected loss.

Answer: Risk-neutral 3-year PD is about 7.2% against about 5.8% real-world. The higher risk-neutral figure reflects risk premia. Use the real-world figure for expected loss and capital, and the risk-neutral figure for pricing and CVA.

Exam tips

Always check whether the question asks for risk-neutral or real-world probabilities. The two answers differ and the options usually include both.
Write the units: bps to decimals, years to match λ. Most lost marks are unit slips.
Know the direction: risk-neutral hazard is typically above real-world, but present it as typical, not guaranteed.
Expect conceptual items too: why reduced-form models fit short-term spreads better than structural models, and why they need market data to calibrate.
For a given spread, a higher recovery assumption raises the implied hazard rate. Be ready to reason about this without calculating.

Practice questions from Introduction to Credit Risk Modeling and Assessment

Reduced-Form Models and Hazard Rates 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 and Hazard Rates: frequently asked questions

What is a hazard rate in a reduced-form credit model?

It is the instantaneous probability of default per unit of time, given that the firm has survived so far. With a constant hazard rate λ, survival to time t is e^(−λt). It is also called the default intensity.

How are credit spreads related to default intensity?

Under risk-neutral pricing, the spread is approximately the hazard rate times loss given default: s ≈ λ × (1 − R). So λ ≈ s ÷ (1 − R). This is an approximation that assumes a flat curve and continuous compounding.

Why is risk-neutral default probability usually higher than real-world?

Market spreads include compensation for risk, not just expected loss. Investors price in default clustering, uncertainty about recovery and illiquidity. So spread-implied probabilities usually exceed historical default frequencies.

How is a reduced-form model different from the Merton model?

Merton links default to the firm's asset value falling below its debt, so default depends on firm structure. A reduced-form model does not model assets. It treats default as a random event with an intensity that is calibrated to market prices.