The assessment of risk in CDOs hinges on two pivotal concepts: probability theory and correlation among underlying assets. While probability theory offers insights into the likelihood of individual default events, correlation provides a measure of how these defaults are linked.

CDOs, structured into senior (*), mezzanine, and equity tranches, each bear varying degrees of risk based on their
priority in loss absorption The risk is inherently tied to the default probabilities of the underlying assets, such as loans or bonds. In this context, the relative frequency of defaults under
different correlation scenarios becomes a focal point of risk assessment.

According to probability theory, the chance of two independent events occurring together is less than the chance of
each occurring separately. In a mathematical context, if P(A) is the probability of event A and P(B) is the probability of event B, then the probability of both occurring simultaneously, assuming
independence, is P(A)*P(B). This principle is crucial for understanding the distribution of risks in CDOs.

For instance, if the probability of event A happening is 0.2 and the probability of event B happening is 0.3, then the
probability of both A and B happening together would be: 0.06 (0.2*0.3).

This probability (0.06) is indeed less than the probability of either event occurring individually (0.2 and 0.3).

Correlation between defaults of various assets in a CDO can significantly influence the overall risk profile. High
correlation implies that defaults are likely to happen in tandem, potentially leading to large-scale simultaneous losses. On the other hand, low correlation suggests that defaults occur more
independently.

In low correlation settings, the defaults of individual assets are largely independent. While the individual risk of
default remains unchanged, the likelihood of multiple assets defaulting at the same time is comparatively lower. This leads to a higher relative frequency of smaller loss events, mostly impacting
the equity or junior tranche. In such scenarios, the senior tranches are shielded from immediate losses due to the dispersion of risk among many independent defaults.

In contrast, a high correlation scenario would mean a lower relative frequency of defaults but with potentially higher
severity. Here, simultaneous defaults can erode the buffers of subordinated tranches quickly, exposing even the senior tranches to significant risk.

(*) Drawing a parallel with societal norms where we prioritize protecting our seniors, in CDO structures, “senior”
tranches are similarly safeguarded, being less subordinated and thus less exposed to initial losses.

Écrire commentaire

0