Merton’s threshold model
About this app
Explore default when firm value falls below a debt threshold and investigate the role of asset correlation.
The model
In the Merton-style threshold model, as used in CreditMetrics, each borrower has a standardised asset return A ∼ N(0, 1). It defaults when A falls below a threshold c; to match the default probability, c = Φ⁻¹(PD). Two borrowers are linked through a common systematic factor M: Aᵢ = √ρ · M + √(1 − ρ) · Zᵢ. The parameter ρ is the asset correlation. It determines how likely both are to default together, and hence the default correlation.
What you see
The first card shows the bell curve of borrower 1’s asset return with the default threshold; the area to its left is exactly PD₁. The second card shows the joint distribution of both returns as a point cloud. The lower-left quadrant, where both returns are below their thresholds, is the joint default. The metrics give P(both default), the product PD₁ · PD₂ under independence, the ratio between them, and the default correlation ρ_D.
Try it
Drag the threshold directly on the card or set PD₁ and PD₂ anywhere from 0.1 % to 50 %. Then raise the asset correlation ρ: the cloud tilts towards the diagonal and more mass lands in the corner quadrant. Compare P(both) with the product of the PDs. Switch off the second borrower to see the single-name model with threshold and PD only.
Frequently asked questions
- How does the Merton model explain credit defaults?
- In Merton’s model a firm defaults when the value of its assets at the horizon falls below the value of its debt. Portfolio models turn this into a threshold on a normally distributed asset return: default if A < Φ⁻¹(PD). Correlation between borrowers comes from shared factors in their asset returns.
- What is the difference between asset correlation and default correlation?
- Asset correlation ρ measures how closely the continuous asset returns of two borrowers move together. Default correlation is the correlation of the default indicators, which are 0 or 1. For small PDs it is much lower than the asset correlation, but it depends strongly on it.
- Why is the probability of a joint default higher than PD₁ · PD₂?
- The product PD₁ · PD₂ only holds if defaults are independent. With positive asset correlation, bad values of the shared factor pull both borrowers down at the same time. The joint default probability rises above the product, and a model that assumes independence understates this concentration risk.
- What can I change in the threshold model app?
- You set PD₁, PD₂ and the asset correlation ρ, drag the threshold on the card and show or hide the second borrower. The app reports P(both default), the product of the PDs, the ratio between them and the default correlation.
Subject: Credit risk
More from Credit risk
- Adverse selection
- Default correlation surface in 3D
- CreditPortfolioView: macro factors
- CreditRisk+: Poisson modelling
- Exposure profile
- Integrated risk measurement
- Rating migration explorer
- PD master scale
A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps