Michael Brenndoerfer | Data & AI, Private Equity, Technology

Michael Brenndoerfer

Credit Risk Fundamentals › Key Parameters

Michael BrenndoerferDecember 21, 20255 min read

Quantitative FinanceAll chapters

Part 10 of 15Risk Management and Financial Risk Controls

Chapter 54Credit Risk Fundamentals › Key Parameters

Reading section 4 of 15 · Key Parameters · 5 minSection 4 of 15 · 5 min

Section 4 of 15 · 5 min read

Credit Risk Fundamentals › Key ParametersOne section · Sections in this chapter4 / 15 ↓

Sections in this chapter15

Reading tools Definitions & inline links

Choose your expertise level to adjust how many terms are explained. Beginners see more tooltips, experts see fewer to maintain reading flow. Hover over underlined terms for instant definitions.

Article links

Make inline references clickable

The key parameters for the market-implied PD calculation are:

CDS Spread: The annual cost of credit protection in basis points. Higher spreads imply higher default risk.
Recovery Rate: The fraction of exposure expected to be recovered in default. Lower recovery rates imply lower default probabilities for a given spread.
Time Horizon: The period over which the default probability is estimated.

Through-the-Cycle versus Point-in-Time PDLink Copied

PD estimation distinguishes between through-the-cycle (TTC) and point-in-time (PIT) approaches, which represent fundamentally different philosophies about how to assess credit risk. Understanding this distinction is essential for interpreting PD estimates correctly and choosing the right approach for each application.

Through-the-Cycle PD

Estimates default probability based on an average of economic conditions over a full credit cycle, producing relatively stable PD values. Used primarily by rating agencies for long-term assessments.

Point-in-Time PD

Reflects current economic conditions and the borrower's present financial state, resulting in more volatile but timely PD estimates. Required for many regulatory and accounting purposes under frameworks like IFRS 9.

These approaches serve different purposes. A TTC assessment asks: "What is the average default probability for this type of borrower across good times and bad?" A PIT assessment asks: "Given what we know about current economic conditions and this specific borrower's current situation, what is the default probability right now?" Both questions are valid, but they serve different purposes.

In[8]:

Code

# Simulating TTC vs PIT PD over an economic cycle
np.random.seed(42)

# Economic cycle indicator (-1 = recession, +1 = expansion)
quarters = np.arange(40)
economic_cycle = np.sin(2 * np.pi * quarters / 20)  # ~5 year cycle

# Base PD for a BBB-rated company
base_pd = 0.02

# TTC remains relatively stable
ttc_pd = np.ones(40) * base_pd * (1 + 0.05 * np.random.randn(40))

# PIT varies with economic conditions
pit_pd = base_pd * (1 - 0.5 * economic_cycle) * (1 + 0.1 * np.random.randn(40))
pit_pd = np.clip(pit_pd, 0.005, 0.10)  # Keep in reasonable bounds

Out[9]:

Visualization

Line chart over ten years with a near-flat Through-the-Cycle PD around 2% and a more variable Point-in-Time PD ranging from roughly 1% to 3.3%; two shaded recession intervals align with PIT increases. — Comparison of Through-the-Cycle and Point-in-Time PD estimates for a BBB-rated company over a simulated 10-year period containing two economic cycles. TTC estimates remain stable while PIT estimates fluctuate with economic conditions.

The choice between TTC and PIT depends on the application. For pricing loans at origination, TTC PD provides stability and ensures consistent treatment of similar borrowers regardless of where we are in the economic cycle. This prevents the procyclical lending behavior where banks extend too much credit during booms (when PIT PDs are low) and restrict credit excessively during downturns (when PIT PDs spike). For calculating accounting provisions or short-term risk management, PIT PD better reflects current conditions and provides a more accurate estimate of near-term expected losses. Basel regulatory frameworks have moved toward PIT estimates for expected credit loss calculations, recognizing that provisions should reflect the best current estimate of losses rather than long-term averages.

Structural Models of DefaultLink Copied

Beyond statistical approaches based on historical default rates or market prices, structural models derived from option pricing theory provide a theoretically grounded framework for PD estimation. These models start from first principles about what causes default and derive default probabilities from observable financial variables. The Merton model, which we'll explore more fully in the next chapter, treats equity as a call option on the firm's assets. Default occurs when asset value falls below the debt's face value at maturity, triggering bankruptcy.

The key insight is that a firm's equity holders have limited liability: if assets fall below debt obligations, they can "put" the firm to creditors and walk away. From the shareholders' perspective, they have the right, but not the obligation, to pay off the debt and keep any residual value. This is precisely the payoff structure of a call option, where the asset value is the underlying and the debt face value is the strike price. This optionality allows us to use option pricing techniques to estimate default probability. Assuming the firm's asset value follows a log-normal distribution (as in the Black-Scholes framework), the probability of default is the probability that the asset value $V$ falls below the debt face value $D$ at maturity $T$ :

\text{PD} = N(-d_2)

where:

$\text{PD}$ : probability of default
$d_2 = \frac{\ln(V/D) + (\mu - \frac{1}{2}\sigma_V^2)T}{\sigma_V\sqrt{T}}$ : distance to default, measuring how many standard deviations the asset value is from the debt face value
$V$ : current asset value
$D$ : face value of debt
$\mu$ : expected asset return
$\sigma_V$ : asset volatility
$T$ : time to maturity
$N(\cdot)$ : cumulative standard normal distribution function

The $d_2$ term, often called the "distance to default," gives an intuitive measure of credit quality. It represents, in standardized units, how far the firm's assets must fall before default becomes inevitable. A higher distance to default indicates a safer credit, while a lower or negative distance to default signals imminent distress. This structural approach has the advantage of linking default probability directly to observable firm characteristics like leverage and asset volatility, and gives economic intuition for why certain firms are riskier than others.