How do I use the hazard rate to calculate cumulative default probability over multiple years?
CFA Level II introduces the hazard rate (conditional default probability) and I'm struggling with converting it to cumulative PD. If the hazard rate is constant at 3% per year, the cumulative PD after 5 years isn't simply 15% — I know it's less. Can someone walk through the math?
The hazard rate framework is elegant and deeply testable at Level II. The key is understanding the relationship between conditional and cumulative probabilities.
Definitions:
- Hazard rate (h): The probability of default in a given year, conditional on having survived to the beginning of that year
- Survival probability: Probability of NOT defaulting through a given period
- Cumulative default probability (PD): Probability of defaulting at any point during the entire period
The Key Relationships:
Survival probability through year t:
S(t) = (1 - h)^t (if hazard rate is constant)
Cumulative default probability:
PD(t) = 1 - S(t) = 1 - (1 - h)^t
Why It's Not Simply h x t:
The hazard rate compounds — each year's default probability applies only to the survivors from the previous year. This is identical to compound interest logic.
Worked Example:
Constant hazard rate h = 3% per year
| Year | Survival at Start | Default in Year | Cumulative PD |
|---|---|---|---|
| 1 | 100.00% | 3.00% | 3.00% |
| 2 | 97.00% | 2.91% | 5.91% |
| 3 | 94.09% | 2.82% | 8.73% |
| 4 | 91.27% | 2.74% | 11.47% |
| 5 | 88.53% | 2.66% | 14.13% |
Calculation Check:
- S(5) = (1 - 0.03)^5 = (0.97)^5 = 0.8587
- PD(5) = 1 - 0.8587 = 14.13% (not 15%!)
The cumulative PD (14.13%) is less than the naive calculation (5 x 3% = 15%) because each year's default probability applies to a shrinking population of survivors.
Variable Hazard Rates:
When the hazard rate changes over time:
S(t) = (1-h_1)(1-h_2)...(1-h_t)
Example: h_1 = 2%, h_2 = 3%, h_3 = 5%
S(3) = (0.98)(0.97)(0.95) = 0.9028
PD(3) = 1 - 0.9028 = 9.72%
Marginal (Unconditional) Default Probability:
The probability of defaulting in year t specifically (not conditional on surviving):
Marginal PD(t) = S(t-1) x h_t
Year 3 marginal PD = 0.98 x 0.97 x 0.05 = 4.75% (not 5%, because some entities already defaulted)
Exam Application — Bond Pricing:
The hazard rate framework connects directly to credit spreads:
Credit spread ≈ h x LGD = h x (1 - Recovery Rate)
If h = 3% and recovery = 40%:
Credit spread ≈ 3% x 60% = 1.80% = 180 bps
This provides a quick sanity check: if a BB bond trades at 250 bps but the hazard rate implies only 180 bps of credit spread, the extra 70 bps reflects liquidity premium.
Exam Tip: Practice computing cumulative PD, survival probabilities, and marginal PD from both constant and variable hazard rates. The connection to credit spreads (spread ≈ h x LGD) is frequently tested.
Master credit modeling in our CFA Level II fixed income modules.
Master Level II with our CFA Course
107 lessons · 200+ hours· Expert instruction
Related Questions
What exactly is the Capital Market Expectations (CME) framework and why does it matter for asset allocation?
How do business cycle phases affect asset class return expectations?
Can someone explain the Grinold–Kroner model step by step with numbers?
How do you forecast fixed-income returns using the building-blocks approach?
PPP vs Interest Rate Parity for forecasting exchange rates — when do I use which?
Join the Discussion
Ask questions and get expert answers.